Downloaded from https://academic.oup.com/jpp/article/56/12/1477/6147469 by Maine Medical Center user on 21 June 2023
`
`lrn"i1
`~ Journalof
`
`Pharmacy and Pharmacology
`
`JPP 2004, 56: 1477–1492
`ß 2004 The Authors
`Received 15 July, 2004
`Accepted 13 August, 2004
`DOI 10.1211/0022357044805
`ISSN 0022-3573
`
`Department of Physiology and
`Pharmacology, University of
`Strathclyde, Strathclyde Institute
`for Biomedical Sciences, Taylor
`Street, Glasgow G4 ONR, UK
`
`Brian Furman, Nigel Pyne
`
`School of Biomedical Sciences,
`University of Ulster, Coleraine,
`BT52 1SA, Northern Ireland
`
`Peter Flatt, Finbarr O’Harte
`
`Correspondence: B. Furman,
`Department of Physiology and
`Pharmacology, University of
`Strathclyde, Strathclyde Institute
`for Biomedical Sciences, Taylor
`Street, Glasgow G4ONR, UK.
`E-mail: b.l.furman@strath.ac.uk
`
`Funding and acknowledgements:
`The study was part funded
`by Diabetes UK, the Research
`and Development Office
`of the Department of Health
`and Personal Social Services
`for Northern Ireland, and
`University of Ulster Research
`Strategy Funding. We are
`grateful to Mrs Pat Owen
`for her assistance.
`
`Review Article
`
`Targeting b-cell cyclic 3050adenosine monophosphate
`for the development of novel drugs for treating
`type 2 diabetes mellitus. A review
`
`Brian Furman, Nigel Pyne, Peter Flatt and Finbarr O’Harte
`
`Abstract
`
`Cyclic 3’5’AMP is an important physiological amplifier of glucose-induced insulin secretion by the
`pancreatic islet

-cell, where it is formed by the activity of adenylyl cyclase, especially in response to
`the incretin hormones GLP-1 (glucagon-like peptide-1) and GIP (glucose-dependent insulinotropic
`peptide). These hormones are secreted from the small intestine during and following a meal, and are
`important in producing a full insulin secretory response to nutrient stimuli. Cyclic AMP influences
`many steps involved in glucose-induced insulin secretion and may be important in regulating
`pancreatic islet

-cell differentiation, growth and survival. Cyclic AMP (cAMP) itself is rapidly
`degraded in the pancreatic islet

-cell by cyclic nucleotide phosphodiesterase (PDE) enzymes. This
`review discusses the possibility of targeting cAMP mechanisms in the treatment of type 2 diabetes
`mellitus, in which insulin release in response to glucose is impaired. This could be achieved by the use
`of GLP-1 or GIP to elevate cAMP in the pancreatic islet

-cell. However, these peptides are normally
`rapidly degraded by dipeptidyl peptidase IV (DPP IV). Thus longer-acting analogues of GLP-1 and
`GIP, resistant to enzymic degradation, and orally active inhibitors of DPP IV have also been devel-
`oped, and these agents were found to improve metabolic control in experimentally diabetic animals
`and in patients with type 2 diabetes. The use of selective inhibitors of type 3 phosphodiesterase
`(PDE3B), which is probably the important pancreatic islet

-cell PDE isoform, would require their
`targeting to the islet

-cell, because inhibition of PDE3B in adipocytes and hepatocytes would induce
`insulin resistance.
`
`Introduction
`
`The prevalence of diabetes mellitus, especially of type 2 diabetes, is increasing and the
`World Health Organisation predicts a world total of > 300 M by 2030. The high
`prevalence of diabetes combined with the associated increased mortality and morbid-
`ity, primarily as a result of macrovascular disease and microvascular long-term com-
`plications, make it a major health problem. Overwhelming evidence, especially from the
`DCCT study (The Diabetes Control and Complications Trial Research Group 1993)
`for type 1 diabetes and the UK Prospective Diabetes Study (Turner 1998) for type 2
`diabetes suggested that good metabolic control would markedly reduce mortality and
`morbidity.
`It is now broadly accepted that type 2 diabetes results from both peripheral insulin
`insensitivity and impaired insulin secretion (Kahn 2003). There is progressive islet

-cell
`failure in patients with type 2 diabetes and a reduction in the

-cell mass (Porte & Kahn
`2001). Thus the islet

-cell remains an important target for the development of drugs for
`treating type 2 diabetes. Drugs are required that both enhance insulin secretion in response
`to normal physiological meal-related nutrient stimuli and that prevent the progressive loss
`of islet

-cell mass. Augmentation of meal-related insulin secretion by drugs should be
`glucose-dependent. This would ensure that increased insulin secretion occurs when
`required, the corollary being that insulin secretion would remain at a basal rate between
`meals, avoiding hyperinsulinaemia and thus hypoglycaemia, as seen with sulphonylureas,
`and the potential cardiovascular risks associated with chronic hyperinsulinaemia and
`insulin resistance (Juhan-Vague & Alessi 1997; Bastard et al 2000). Physiological,
`meal-related insulin secretion occurs in response to absorbed nutrients, primarily glucose.
`Glucose stimulates insulin secretion following its transport into and metabolism in the
`
`1477
`
`Novo Nordisk Exhibit 2004
`Mylan Pharms. Inc.v. Novo Nordisk A/S
`IPR2023-00722
`Page 00001
`
`

`

`Downloaded from https://academic.oup.com/jpp/article/56/12/1477/6147469 by Maine Medical Center user on 21 June 2023
`
`i ~ ~
`CY---.
`ll~ J
`
`, ,._
`
`l '
`
`I
`
`'\
`
`I v
`
`\~ _ -
`
`8p
`
`1478
`
`Brian Furman et al
`
`glucose
`
`K+
`
`K+
`
`glucose
`
`ATP
`
`nitric oxide
`
`+
`
`depolarization
`
`Ca2+
`stores
`
`+
`
`Ca2+
`
`GLP-1
`
`Ca2+
`
`GLP-1
`receptor
`
`adenylyl cyclase
`
`GIP
`
`GIP
`receptor
`
`guanylyl cyclase
`
`metabolism
`
`GDP
`
`GTP
`
`Transcription
`
`Rap1
`
`Rap1
`
`GTP
`
`Cyclic 3’5’GMP
`
`PK A
`
`PDE1
`+
`
`PDE5
`
`−
`
`Ca2+
`
`5’GMP
`
`Epac
`
`+
`
`+
`
`+
`
`Cyclic 3’5’AMP
`
`+
`
`PDE1
`
`PDE3B
`
`+
`
`+
`
`+
`
`5’AMP
`
`+
`exocytosis
`
`IGF-1
`receptor
`
`leptin
`receptor
`
`Insulin
`receptor
`
`insulin
`
`Ca2+
`
`IGF-1
`
`liver
`
`adipose
`tissue
`
`leptin
`
`
`
`Figure 1 Schematic diagram of the pancreatic islet

-cell showing the sites of action (indicated by the star symbol) of cyclic 3050AMP in
`modulating glucose-induced insulin secretion and other

-cell activities. The pathways for the formation and destruction of cAMP and the
`hormonal factors influencing these pathways are shown also.
`
`

-cell, generating ATP which closes KATP channels, resulting
`in depolarization and calcium influx (Figure 1). The effect of
`glucose on insulin secretion can be amplified by signalling
`pathways involving inositol trisphosphate and diacylglycerol
`derived from activation of phospholipase C (Howell et al
`1994; Gilon & Henquin 2001) and by cyclic 3050AMP follow-
`ing activation of adenylyl cyclase (Howell et al 1994).
`Glucose itself has long been known to elevate the pancreatic
`islet

-cell cAMP content (Grill & Cerasi 1973). Although
`glucose-induced insulin secretion does not appear to require
`cAMP or the protein kinase A (PKA)-system (Grill & Cerasi
`1973; Sharp 1979; Persaud et al 1990; Lester et al 1997),
`increases in cAMP augment the direct effect of glucose.
`This cyclic nucleotide is generally accepted as an important
`amplifier of glucose-induced insulin release (Holz & Habener
`1992), particularly when its levels are increased by glucose
`itself (Harnda¨ hl et al 2002) and, importantly, by various gut
`hormones implicated as incretins, which are secreted from
`the small intestine in response to the presence of glucose and
`other nutrients in the gut. It has also been suggested that
`cAMP is a competence factor for normal islet

-cell respon-
`siveness to glucose (Schuit 1996; Huypens et al 2000).
`Numerous hormones that stimulate insulin secretion increase
`islet

-cell cAMP, including glucagon (Huypens et al 2000),
`ACTH (Al-Majed et al 2004), and pituitary adenylate
`cyclase-activating polypeptide (PACAP; Filipsson et al
`(2001)). On the other hand, hormones inhibiting insulin
`secretion (galanin (Drews et al 1994), adrenaline (Peterhoff
`et al 2003)), reduce cAMP, although their effects on cAMP
`
`may not necessarily explain the inhibition of secretion. The
`major incretins are glucose-dependent insulinotropic poly-
`peptide (GIP) secreted by intestinal K-cells and glucagon-like
`peptide 1 secreted by intestinal L-cells. These hormones aug-
`ment glucose-induced insulin secretion through activating
`adenylyl cyclase, leading to an increase in islet

-cell cAMP.
`Their importance is illustrated clearly by the marked glucose
`intolerance and impairment of insulin secretion seen in mice
`lacking receptors for both hormones (Preitner et al 2004).
`
`Secretion and actions of glucagon-like peptide-1
`(GLP-1)
`
`GLP-1 is derived from the processing of the preprogluca-
`gon gene in intestinal L-cells (Lund et al 1982). Various
`products of the post-translational processing of progluca-
`gon exist including GLP-1(7–36)amide and GLP-1(7–37),
`which are equipotent stimulators of glucose-dependent
`insulin secretion (Holst et al 1987; Kreymann et al 1987;
`Mojsov et al 1987). For the purpose of this review, GLP-1
`refers to the active form GLP-1(7–36)amide.
`GLP-1 is secreted from intestinal L-cells in response to
`the absorption of glucose, other sugars, fatty acids and to a
`minor extent amino acids (MacIntosh et al 2001; Feinle et al
`2002; Brubaker & Anini 2003). The GLP-1 secretory
`actions of metabolizable nutrients may be linked to cellular
`ATP generation, KATP channel blockade and elevation
`of intracellular Ca2þ
`(Gribble & Reimann 2002). There is
`also a neurally-mediated arm to GLP-1 secretion and a
`
`Novo Nordisk Exhibit 2004
`Mylan Pharms. Inc.v. Novo Nordisk A/S
`IPR2023-00722
`Page 00002
`
`

`

`Downloaded from https://academic.oup.com/jpp/article/56/12/1477/6147469 by Maine Medical Center user on 21 June 2023
`
`stimulatory effect of GIP. These pathways may be espe-
`cially important for release of GLP-1 from the distal region
`of the small intestine (Holst et al 1987; Filipsson et al 2000).
`The insulinotropic actions of GLP-1 are evident in islets
`(Komatsu et al 1989; Weir et al 1989) in normal and diabetic
`animals (Hendrick et al 1993; Scrocchi et al 1998; O’Harte
`et al 2000a, 2000b, 2001) and in healthy and diabetic volun-
`teers (Gutniak et al 1992).
`Other physiological actions of GLP-1 target the pancrea-
`tic islets, liver, adipose tissue and skeletal muscle. In pan-
`creatic islets, GLP-1 stimulates insulin and somatostatin
`secretion (Holst et al 1987; Mojsov et al 1987; Eissele et al
`1990; Creutzfeldt et al 1996) and inhibits glucagon secretion
`(Creutzfeldt et al 1996). GLP-1 stimulates insulin gene tran-
`scription (Drucker et al 1987; Fehmann & Habener 1992;
`Skoglund et al 2000), pancreatic islet cell proliferation
`(Buteau et al 1999; Perfetti et al 2000), and

-cell replication
`(Edvell & Lindstrom 1999). GLP-1 leads to the differentia-
`tion of pancreatic ductal AR42J cells into glucagon- and
`insulin-producing cells (Zhou et al 1999), through a PDX-1
`dependent pathway (De La Tour et al 2001; Hui et al 2001).
`Lesser known actions of GLP-1 include promotion of glu-
`cose uptake and glycogen formation in liver and skeletal
`muscle (Valverde et al 1994; Villanueva-Penacarrillo et al
`1994; O’Harte et al 1997). GLP-1 also stimulates lactate
`production, glucose uptake and glycogen storage in skeletal
`muscle (O’Harte et al 1997).
`
`Secretion and actions of GIP
`
`The 42 amino acid peptide GIP was isolated from a crude
`porcine CCK extract and shown to inhibit histamine-
`induced gastric acid secretion (Brown et al 1969). How-
`ever, its major physiological role as a glucose-dependent
`stimulator of insulin secretion is now well recognized
`(Gault et al 2003a).
`GIP is secreted from intestinal K-cells by glucose and
`other actively transported sugars and by fatty acids.
`Amino acids however are only a weak stimulus.
`Although GIP stimulus-secretion coupling pathways are
`poorly understood, there appears to be a close link with
`K-cell metabolism (Tseng et al 1994). Indeed it appears
`likely that signal transduction of intestinal K- and L-cells
`will be found to have many parallels to other glucose-
`sensitive secretory cells, most notably the pancreatic

-
`cell (Purrello & Rabuazzo 2000). Notably, insulin forms
`an effective feedback loop by inhibiting nutrient stimu-
`lated GIP secretion (Bryer-Ash et al 1994).
`GIP is a potent glucose-dependent stimulator of insulin
`secretion in all in-vitro and in-vivo systems tested (Pederson
`et al 1998a, 1998b; Drucker 2003; Vilsboll & Holst 2004).
`Additional actions on the

-cell include stimulation of
`proinsulin gene transcription and translation (Fehmann &
`Go¨ ke 1995; Wang et al 1996), enhancement of the growth,
`differentiation and survival of pancreatic

-cells (Trumper
`et al 2001; Pospisilik et al 2003). Extra pancreatic actions of
`GIP appear to enhance its glucose-lowering ability by inhi-
`biting hepatic glucose production (Elahi et al 1986) and
`promoting glucose uptake in isolated mouse diaphragm
`muscle (O’Harte et al 1997). Functional GIP receptors
`
`Targeting cAMP in the islet

-cell
`
`1479
`
`have also been identified on adipocytes (Yip et al 1998),
`where GIP has been shown to stimulate glucose transport
`(Eckel et al 1979), fatty acid synthesis (Oben et al 1991) and
`lipoprotein lipase activity (Knapper et al 1995).
`
`Cyclic AMP in the pancreatic islet

-cell
`GLP-1 and GIP exert their insulinotropic actions through
`activation of G-protein coupled receptors
`(GPCRs)
`(Thorens 1992; Usdin et al 1993). This results in activation
`of adenylyl cyclase and an increase in intracellular cAMP.
`Their effects in stimulating insulin synthesis and in the
`enhancement of the growth, differentiation and survival
`of pancreatic

-cells are probably also mediated by
`cAMP. However, there is evidence that other pathways,
`such as the transactivation of epidermal growth factor
`receptors, may be involved in stimulation of

-cell proli-
`feration (Buteau et al 2003). Cyclic AMP has many effects
`within the

-cell, as described below (see also Figure 1).
`
`Insulin secretion Several mechanisms may mediate the
`effect of cAMP in augmenting glucose-induced insulin
`secretion. These include increased opening of voltage-sensi-
`tive Ca2þ
`channels (Kanno et al 1998), calcium-induced
`Ca2þ
`release (Kang et al 2001), activation of ryanodine
`receptors in the endoplasmic reticulum (Islam et al 1998;
`Holz et al 1999), stimulation of

-cell lipolysis (Yaney et al
`2001) and direct effects on exocytosis (Harnda¨ hl et al 2002).
`Most actions of cAMP in the

-cell appear to be mediated
`through PKA-catalysed phosphorylation events. There is
`evidence for specific targeting of PKA to particular sub-
`cellular locations within the

-cell by A-kinase anchoring
`proteins (AKAPs) and disruption of AKAPs inhibits, for
`example, the effects of cAMP in increasing intracellular
`Ca2þ
`and stimulating insulin secretion (Lester et al 1997).
`However, some effects of the cyclic nucleotide on exocytosis
`are partly PKA-independent (Renstrom et al 1997). PKA-
`independent effects on exocytosis may be mediated by the
`cAMP-binding proteins known either as cAMP-regulated
`guanine nucleotide exchange factors (GEFs) or exchange
`proteins activated by cAMP (Epacs) which target the small
`G-protein Rap1 (Kopperud et al 2003). The pancreatic islet
`

-cell expresses both Epac 1 and Epac 2 (Holz 2004).
`Treatment of pancreatic islets with antisense oligodeoxy-
`nucleotides against Epac reduced the effect of a permeant
`cAMP analogue in augmenting glucose-induced insulin
`secretion (Kashima et al 2001). A recently described, novel
`cAMP analogue, 8-(4-chloro-phenylthio)-20-O-methylade-
`nosine-30-50-cyclic monophosphate
`(8-pCPT-20-O-Me-
`cAMP) activated Epac but not PKA. This analogue was
`shown to mobilize Ca2þ
`from intracellular Ca2þ
`stores via
`Epac-mediated Ca2þ
`-induced Ca2þ
`release in human pan-
`creatic

-cells and INS-1 insulin-secreting cells (Kang et al
`2003). The cAMP-mediated insulinotropic action of the
`incretin factors glucose-dependent insulinotropic peptide
`(GIP) and glucagon-like peptide 1 (GLP-1) (Drucker et al
`1987) are mediated in part through PKA, but PKA-inde-
`pendent actions have also been demonstrated and probably
`involve Epac, as evidenced by the effects of antisense oligo-
`deoxynucleotides against Epac (Kashima et al 2001).
`
`Novo Nordisk Exhibit 2004
`Mylan Pharms. Inc.v. Novo Nordisk A/S
`IPR2023-00722
`Page 00003
`
`

`

`Downloaded from https://academic.oup.com/jpp/article/56/12/1477/6147469 by Maine Medical Center user on 21 June 2023
`
`1480
`
`Brian Furman et al
`
`Other roles in the

-cell Glucose and the insulinotropic
`hormones have actions on the

-cell which extend beyond
`the acute stimulation of insulin secretion. These include
`increased insulin synthesis and the modulation of

-cell
`growth, differentiation and apoptosis. In addition to its
`role in amplifying glucose-induced insulin secretion,
`cAMP may mediate or modulate these other effects.
`Glucose-mediated increases in insulin synthesis involve the
`phosphorylation of the transcription factor pancreatic duo-
`denal homeobox-1 (PDX-1) and its translocation to the
`nucleus (Elrick & Docherty 2001). While the effects of
`glucose itself on PDX-1 are not mediated by cAMP, there
`is strong evidence for the importance of cAMP in GLP-1-
`dependent stimulation of PDX-1 in the

-cell, as well as its
`translocation to the nucleus and its activation of the insulin
`gene promoter (Wang et al 2001). However, the role of
`cAMP in regulating insulin synthesis is unclear since the
`adenylyl cyclase activator forskolin or the cAMP analogue
`8-bromo-cAMP suppressed insulin transcription in INS-1
`cells in a PKA-independent manner (Ding et al 2003).
`Cyclic AMP may mediate effects of glucose in stimulating
`the expression of immediate early response genes such as
`c-myc (Jonas et al 2001) and c-fos (Susini et al 1998). In this
`context, cAMP can either activate or inhibit the p42/p44
`mitogen-activated protein kinase pathway depending on the
`cell type and conditions (Calleja et al 1997). Glucose itself
`activates the p42/p44 MAPK pathway (Benes et al 1998,
`1999) and this effect is amplified, although not mediated by
`cAMP. Indeed, the activation of this pathway by GIP is
`cAMP and PKA-dependent (Ehses et al 2002). Regulation
`of gene expression by cAMP may have relevance to effects
`on

-cell growth, differentiation and apoptosis. For exam-
`ple, Andersen et al (1996) showed that elevated cAMP levels
`protected rat islets against interleukin-1

-mediated induc-
`tion of nitric oxide synthase and nitric oxide production.
`Excessive nitric oxide production appears to mediate sub-
`sequent apoptosis of

-cells. On the other hand,

-cell lines
`were made more susceptible to apoptosis following expo-
`sure to dibutyryl cAMP (Loweth et al 1996) or the cAMP
`elevating agent forskolin (Ahmad et al 2000b).
`
`Formation of cAMP in the

-cell
`The pancreatic islet

-cell was shown to express several of
`the nine adenylyl cyclases (AC), including AC1, AC2,
`AC3, AC4, AC5, AC6 and AC8 (Leech et al 1999;
`Guenifi et al 2000; Delmeire et al 2003),
`including the
`calcium-calmodulin activated AC1, AC3 and AC8.
`
`Destruction of cyclic nucleotides in pancreatic
`islet b-cell
`
`The hydrolysis of cAMP and cyclic GMP (cGMP) to their
`biologically inactive 50 derivatives is catalysed by the family
`of enzymes known as the cyclic nucleotide phosphodi-
`esterases, which provide the only known system for destroy-
`ing these cyclic nucleotides. Currently there are 11 known
`gene families of CN-PDEs (PDE1–PDE11), comprising
`more than 50 enzymes with differences in their substrate
`selectivity (cAMP vs cGMP), kinetics, allosteric regulation,
`
`tissue distribution and susceptibility to pharmacological
`inhibition (reviewed by Perry & Higgs 1998; Soderling &
`Beavo 2000; Mehats et al 2002). Table 1 shows the sub-
`strate, the Km values, particular properties and inhibitors
`of some of these enzymes, particularly those which have
`been identified in pancreatic islets. In view of the clear
`importance of cyclic nucleotides, especially cAMP, in the
`pancreatic islet

-cell, it is important to understand the
`nature and the roles of the phosphodiesterase (PDE)
`enzymes expressed in pancreatic islet

-cells.
`
`PDE isoforms present in the islets
`
`Most studies to elucidate the PDE isoforms present in the
`islet

-cell have been undertaken using pancreatic islets,
`which contain four endocrine cell types; these comprise
`the insulin secreting

-cells, the glucagon secreting A cells,
`the somatostatin secreting D cells and the pancreatic poly-
`peptide (PP) cells. Non-endocrine cells, including blood
`vessels are also present. To look more selectively at the islet
`

-cell, insulin-secreting cell lines have been used extensively,
`although these have limitations as models for native

-cells.
`Evidence was obtained for the presence of PDE1 (calcium-
`calmodulin activated), PDE3 (cyclic 3050 GMP-inhibited)
`and PDE4 (cAMP-specific) in islets and

-cell lines.
`There is good evidence for the presence of PDE3 in
`insulin secreting cells for the role of this isoform in regulat-
`ing the cAMP pool relevant to insulin secretion. Thus
`membrane and cytosolic fractions of islet homogenates
`contained a low Km (1.4–2.2 M) cAMP PDE activity
`(Shafiee-Nick 1995) that was inhibited by 60–70% using
`10 M cGMP (Furman & Pyne 1990). The pellet fraction of
`homogenates of BRIN BD11 cells was also found to con-
`tain a cAMP PDE activity that was potently (IC50 0.7 M)
`inhibited (up to 30–40%) by cGMP (Ahmad et al 2000b).
`Of the two isoforms of type 3 phosphodiesterase (PDE3),
`PDE3A and PDE3B, only PDE3B appears to be expressed
`in the

-cell. Western blotting of extracts of rat pancreatic
`islets and the

-cell line HIT-T15, using a polyclonal anti-
`body against a GST-PDE3B, revealed a single protein band
`corresponding in size to the PDE3B protein found in
`extracts of rat epididymal adipose tissue (Zhao et al 1997).
`Immunostaining of rat islets showed that PDE3B was
`expressed only in cells that were co-stained with anti-insulin
`antibodies (Zhao et al 1997). Reverse transcription-poly-
`merase chain reaction (RT-PCR) using BRIN-BD11 cell
`RNA and PDE3B-specific primers demonstrated a product
`showing > 97% sequence homology with rat adipose tissue
`PDE3B (Ahmad et al 2000a, b).
`The selective PDE3 inhibitors SK&F 94836 and Org
`9935 potently inhibited rat islet PDE activity, especially in
`membrane fractions (Shafiee-Nick et al 1995) with up to
`85% of membrane-bound PDE being inhibited. Similar
`observations were made in human islets (Shafiee-Nick
`et al 1994) and in the BRIN BD11 insulin secreting cell
`line (Ahmad et al 2000a), although in homogenates of
`this cell line, unlike in islets, the drugs inhibited cAMP
`hydrolysis only in the membrane fractions. Rat and
`human islets were shown by Parker et al (1995) to contain
`a milrinone-sensitive PDE, accounting for up to 70% of
`
`Novo Nordisk Exhibit 2004
`Mylan Pharms. Inc.v. Novo Nordisk A/S
`IPR2023-00722
`Page 00004
`
`

`

`Downloaded from https://academic.oup.com/jpp/article/56/12/1477/6147469 by Maine Medical Center user on 21 June 2023
`
`Targeting cAMP in the islet

-cell
`
`1481
`
`Table 1 Properties of phosphodiesterases that have been identified in the pancreatic islet

-cell
`
`Enzyme Reference
`
`Substrate Km (mM) Properties/comment
`
`Inhibitors
`
`PDE1A Han et al (1999)
`PDE1B Clapham & Widerspin (2001)
`PDE1C Yan et al (1995)
`
`PDE3A Perry & Higgs (1998),
`Mehats et al (2002)
`PDE3B Pyne et al (1987),
`Harrison et al (1986)
`
`cGMP
`cAMP
`
`3
`1–30
`
`cAMP
`
`0.5
`
`Activated by Ca/calmodulin.
`Evidence for PDE1C in
`insulin secreting cells
`
`Zaprinast (1/5), vinpocetine (1A)
`SCH51866 (also 5,9,10A),
`8-MM-IBMX (1C)
`
`Inhibited by cGMP
`Phosphorylation by PKA and
`insulin-sensitive kinase.
`Evidence for PDE3B but not
`PDE3A in insulin secreting cells
`
`Milrinone, Org 9935, siguazodan,
`cilostamide, enoximone, OPC3911.
`No evidence that any
`of these is selective for PDE3B
`vs PDE3A
`
`Torphy (1998), Houslay (1998) cAMP
`
`0.2–4
`
`PDE4A
`PDE4B
`PDE4C
`PDE4D
`
`Phosphorylation by PKA and
`p42/p44 MAPK (PDE4D3).
`There is no indication which
`subtype(s) present in insulin
`secreting cells.
`
`PDE5
`
`Corbin & Francis (1999)
`
`cGMP
`
`2
`
`Phosphorylation by PKA/PKG
`
`PDE6
`
`Pittler et al (1990)
`
`cGMP
`
`60–100 Has been identified recently in
`pancreatic

-cells (Kaminski &
`Morgan 2004)
`
`Rolipram, RP73401, zardaverine,
`CP80633, CDP840, LAS31025,
`SB207499
`
`Zaprinast, sildenafil, dipyridamole,
`SCH51866 (also 1,9, 10A)
`
`Zaprinast, dipyridamole
`
`PDE8A,
`PDE8B
`
`Fisher et al (1998)
`
`cAMP
`
`0.7
`
`Dipyridamole
`
`IBMX-insensitive. The only
`evidence for its presence in insulin
`secreting cells is the observation
`that approximately 10% of the
`cAMP PDE activity is insensitive
`to IBMX.
`
`PDE10A Soderling & Beavo (2000)
`
`cAMP
`cGMP
`
`0.05
`3
`
`SCH51866 (also 1,5,9) dipyridamole
`
`total islet PDE activity. The PDE3 activity found in the
`soluble fraction of homogenates of rat islets might be
`PDE3A from blood vessels or other non-

-cell tissue,
`since PDE3A was shown exclusively in the cytosolic frac-
`tions of all tissues studied, whereas PDE3B was expressed
`in particulate fractions (Liu & Maurice 1998).
`Other isoforms also contribute to islet

-cell PDE
`activity. Both islets (Sugden & Ashcroft 1981; Lipson &
`Oldham 1983; Capito et al 1986) and the

-cell lines BRIN
`BD11 (Ahmad et al 2000a) and

-TC3 (Han et al
`1999) contain a Ca-calmodulin activated PDE. The
`PDE1/PDE5 inhibitor zaprinast produced a 14–30% inhi-
`bition in membrane and cytosolic fractions of rat islet
`homogenates (Shafiee-Nick et al 1995) and 25% inhibition
`of cAMP/cGMP PDE activity in BRIN BD11 cells
`(IC501–5 M) (Ahmad et al 2000a). Han et al (1999)
`showed that the presence of a PDE that was inhibited by
`zaprinast (IC50 4.5 M) and 8-MM-IBMX (IC50 7.5 M)
`but not by vinpocetine (IC50 > 100 M for PDE1C vs 1.4
`and 9.8 M for PDE1A and PDE1B, respectively). These
`findings were compatible with the presence of PDE1C but
`not PDE1A or PDE1B and this was supported by RT-
`PCR (Han et al 1999).
`A role for the cAMP-selective PDE4 in the hydrolysis
`of

-cell cAMP was supported by the inhibition of islet
`(Shafiee-Nick et al 1995; Parker et al 1995) or BRIN-
`BD11 cell (Ahmad et al 2000a) PDE by the selective
`PDE4 inhibitor rolipram. Han et al (1999) demonstrated
`
`the expression of PDE4A and PDE4D in

TC3 cells using
`RT-PCR.
`
`The importance of PD3B in regulating insulin
`secretion
`
`studies, using biochemical and pharmacological
`Our
`approaches, first suggested the role of PDE3 in modulating
`glucose-induced insulin secretion in islets of rat (Furman &
`Pyne 1990; Shafiee-Nick et al 1995) and man (Shafiee-Nick
`et al 1994). This has been generally confirmed by others in
`islets (Parker et al 1995; Zhao et al 1997; Harnda¨ hl et al
`2002), and by ourselves and others in insulin secreting cell
`lines (Ahmad et al 2000a; Harnda¨ hl et al 2002). The impor-
`tance of PDE3B in regulating

-cell cAMP in the context of
`insulin secretion was demonstrated by adenovirus-mediated
`over-expression of PDE3B in a

-cell
`line or in islets
`(Harnda¨ hl et al 2002) and by using transgenic animals
`over-expressing PDE3B in the

-cell (Harnda¨ hl et al 2004).
`Those studies clearly showed that glucose-induced, as well
`as GLP-1-induced insulin secretion was impaired by
`PDE3B over-expression in-vitro and in-vivo.
`Although there is very strong evidence supporting the
`role of PDE3B in the

-cell, some studies have suggested
`the importance of other PDEs. Rolipram, a selective PDE4
`inhibitor, augmented glucose-induced insulin secretion in
`

-cell lines (Han et al 1999; Ahmad et al 2000a), although
`not in islets (Shafiee-Nick et al 1995). This discrepancy
`
`Novo Nordisk Exhibit 2004
`Mylan Pharms. Inc.v. Novo Nordisk A/S
`IPR2023-00722
`Page 00005
`
`

`

`Downloaded from https://academic.oup.com/jpp/article/56/12/1477/6147469 by Maine Medical Center user on 21 June 2023
`
`1482
`
`Brian Furman et al
`
`between islets and cell lines may be due to differences in the
`compartmentalization of PDE4 isoforms, allowing PDE4
`to interact with relevant cAMP pools in transformed cell
`lines but not in native

-cells. A role for PDE1 in both the
`

-TC3 cell line and in islets was suggested by Han et al
`(1999) who showed that 8-MM-IBMX, purported to be a
`selective inhibitor of the Ca-calmodulin activated PDE1C
`isoform, augmented glucose-induced insulin secretion. The
`importance of PDE1 in regulating

-cell cAMP remains to
`be determined.
`
`Exploitation of cAMP mechanisms in the
`development of drugs to treat type 2 diabetes
`
`From the foregoing it is clear that increasing

-cell levels
`of cAMP will augment glucose-induced insulin secretion.
`Therapeutically, this could be achieved by activating

-
`cell adenylyl cyclase using analogues of GLP-1 or GIP,
`increasing the availability of endogenous GLP-1 and GIP
`by preventing their normal very rapid breakdown or by
`preventing the breakdown of cAMP in the

-cell using
`inhibitors of phosphodiesterase selective for the most
`important

-cell isoform i.e. PDE3B.
`
`GLP-1 analogues
`
`The metabolic stability and biological activity of many
`GLP-1 analogues have been investigated (Deacon et al
`1998; Knudsen et al 2000; Xiao et al 2001; Green et al
`2003a, b, 2004a). So far, limited clinical trials have been
`conducted using GLP-1 (Gutniak et al 1992; Meier et al
`2003; Meneilly et al 2003), GLP-1 analogues (Rolin et al
`2002; Chang et al 2003; Holz & Chepurny 2003) and
`exendin-4 (Egan et al 2002, 2003). However,
`clear
`improvements in glucose tolerance have been observed in
`experimental animals (Holst 1999).
`
`N-terminal (His7 modified) analogues of GLP-1
`Substitution of His7 with several other amino acids indi-
`cated that the imidizole ring of His7 was crucial for GLP-1
`action (Hareter et al 1997). It was shown that N-terminal
`extension of GLP-1, with insertion of an acetyl group at
`the -amino group of His7, did not significantly affect
`GLP-1 action. N-terminally extended peptides with imidi-
`zole-lactic acid, N-methyl, and -methyl groups at His7
`(Table 2) are more resistant to dipeptidyl peptidase IV
`(DPP IV) than native GLP-1, but exhibit reduced affinity
`for the GLP-1 receptor and compromised ability to sti-
`mulate cAMP production (Hareter et al 1997). His7-gluci-
`tol-GLP-1 was shown to be resistant to DPP IV whilst
`maintaining antihyperglycaemic activity in-vivo (Table 2)
`(O’Harte et al 2000a, b). Most recently, two additional
`analogues, N-acetyl-GLP-1 and N-pyroglutamyl-GLP-1,
`have been shown to be completely resistant to DPP IV and
`human plasma degradation, as well as exhibiting potent
`receptor binding, cAMP production and insulin secretory
`activity in-vitro (Table 2). In obese diabetic (ob/ob) mice,
`N-acetyl-GLP-1 and N-pyroglutamyl-GLP-1 displayed
`potent insulinotropic actions, with N-pyroglutamyl-GLP-1
`
`being particularly effective compared with the antihy-
`perglycaemic effects of the native peptide (Green et al
`2004b).
`
`Position 2 (Ala8 substituted) analogues of GLP-1
`Substitution of Ala8 for Gly conferred increased resistance
`to DPP IV and, despite a reduced affinity for the GLP-1
`receptor, corrected the fasting hyperglycaemia and glucose
`intolerance of diabetic mice (Burcelin et al 1999) (Table 2).
`Similarly, substitution with Ser significantly increased the
`plasma stability of GLP-1 without impairing its insulino-
`tropic activity in rats (Ritzel et al 1998)
`(Table 2).
`Substitution of Ala8 for Thr, Gly, Ser and -aminoiso-
`butyric acid (Aib) significantly prolonged biological half-
`lives in-vivo compared with native GLP-1, but nevertheless
`bound to the GLP-1 receptor with high affinities (Deacon
`et al 1998) (Table 2). However, only native GLP-1 and
`(Aib8)GLP-1 significantly improved insulin output over
`basal conditions (Deacon et al 1998). Similarly, analogues
`with D-Alanine (D-Ala), Ser and Gly substitutions at posi-
`tion 8 were substantially more resistant to DPP IV than
`native GLP-1, and had similar or enhanced biological half-
`life and potency (Siegel et al 1999a, b) (Table 2). Substitution
`of Ala8 for Gly and aminohexanoic acid (Aha) (Table 2)
`produced analogues that stimulated insulin secretion and
`intracellular cAMP to a similar degree to native GLP-1
`(Doyle et al 2001). In-vivo they lowered circulating blood
`glucose and increased blood insulin concentrations in
`Zucker rats. More recently, Green et al (2003b) character-
`ized two novel Ala8-substituted analogues of GLP-1,
`namely (Abu8)GLP-1 and (Val8)GLP-1. These were
`shown to be completely resistant to the actions of DPP IV
`or human plasma (Table 2). Receptor binding studies
`demonstrated that (Abu8)GLP-1 and (Val8)GLP-1 bound
`the GLP-1 receptor with high affinity, but that this was
`reduced compared with native GLP-1 (Table 2). Although
`active stimulators of
`intracellular cAMP production,
`(Abu8)GLP-1 and (Val8)GLP-1 were 1.5- and 3.5-fold less
`potent, respectively, than native GLP-1 (Green et al 2003b).
`Despite these losses in receptor affinity and cAMP produc-
`tion, this did not compromise insulinotropic activity either
`in-vitro or in-vivo (Table 2). This may be partially explained
`by the diverse mechanisms of action of GLP-1 on the
`pancreatic

-ce