WO1997046669A1 - Transgenic mammals lacking expression of amylin - Google Patents

Abstract

Transgenic non-human animals, including mice, with suppressed amylin gene activity, and uses therefor.

Description

TRANSGENIC MAMMALS LACKING EXPRESSION OF AMYLIN

This application is a continuation-in-part of U.S. application Serial No. 08/660,134 filed June 7, 1996.

Field of the Invention

This invention relates to mammals in which the expression of the amylin gene has been suppressed. More specifically, the invention concerns transgenic mammals with decreased or completely suppressed expression of the endogenous amylin gene.

Background Amylin

The structure and biology of amylin have previ¬ ously been reviewed. See, for example, Rink et al. , Trends in Pharmaceutical Sciences , 14:113-118 (1993) ; Gaeta and Rink, Med. Chem. Res . , 3:483-490 (1994) ; and, Pittner et al . , J. Cell . Biochem . , 55S:19-28 (1994) . Amylin is a 37 amino acid protein hormone. It was iso¬ lated, purified and chemically characterized as the major component of amyloid deposits in the islets of pancreases of deceased human Type 2 diabetics (Cooper et al . , Proc. Na tl . Acad. Sci . USA, 84:8628-8632 (1987)) . The amylin molecule has two important post-translational modifica¬ tions: the C-terminus is amidated, and the cysteines in positions 2 and 7 are cross-linked to form an N-terminal loop (disulphide linkage) . The sequence of the open reading frame of the human amylin gene shows the presence of a Lys-Arg dibasic amino acid proteolytic cleavage site, prior to the N-terminal codon for Lys, and a Gly prior to the Lys-Arg proteolytic signal at the C-terminal position, a typical sequence for amidation by protein amidating enzyme, PAM (Cooper et al. , Biochem . Biophys . Acta , 1014:247-258 (1989)) . Amylin is the subject of United States Patent No. 5,367,052, issued November 22, 1995.

In Type 1 diabetes, amylin has been shown to be deficient and combined replacement with insulin has been proposed as a preferred treatment over insulin alone in all forms of diabetes. The use of amylin and other amylin agonists for the treatment of diabetes mellitus is the subject of United States Patent No. 5,175,145, issued December 29, 1992. Pharmaceutical compositions containing amylin and amylin plus insulin are described in United States Patent No. 5,124,314, issued June 23, 1992.

Excess amylin action has been said to mimic key features of Type 2 diabetes and amylin blockade has been proposed as a novel therapeutic strategy. It has been disclosed in United States Patent No. 5,266,561, issued November 30, 1993, that amylin causes reduction in both basal and insulin-stimulated incorporation of labeled glucose into glycogen in skeletal muscle. The latter effect was also disclosed to be shared by calcitonin gene related peptide (CGRP) (see also Leighton and Cooper, Nature, 335:632-635 (1988)) . Amylin and CGRP were approx¬ imately equipotent, showing marked activity at 1 to 10 nM. Amylin is also reported to reduce insulin-stimulated uptake of glucose into skeletal muscle and reduce glycogen content (Young et_. al . , Arner. J. Physiol . , 259:457-46-1 (1990) ) . The treatment of Type 2 diabetes and insu¬ lin resistance with amylin antagonists is disclosed.

Both the chemical structure and the gene se- quence of amylin have been said to support the determina¬ tion that it is a biologically active or "messenger" molecule. The chemical structure is nearly 50% identical to the CGRPs, also 37 amino acid proteins which are widespread neurotransmitters with many potent-biological actions, including vasodilation. Amylin and CGRP share the ²Cys-⁷Cys disulphide bridge and the C-terminal amide, both of which are essential for full biologic activity (Cooper et al. , Proc. Natl . Acad. Sci . USA, 85:7763-7766

(1988) ) . Amylin reportedly may be one member of a family of related peptides which includes CGRP, insulin, insulin- like growth factors and the relaxins and which share common genetic heritage (Cooper et_. al. , Prog. Growth Factor Research, 1:99-105 (1989)) . The two peptides calcitonin and CGRP-1 share common parentage in the calcitonin gene where alternative processing of the primary RNA transcript leads to the generation of the two distinct peptides, which share only limited sequence homology (about 30%) (Amara et al . , Science, 229:1094-1097 (1985) ) . The amylin gene sequence is typical for a secreted messenger protein, with the mRNA coding a prepro- peptide with processing sites for production of the secreted protein within the Golgi or secretary granules. Amylin is mainly co-localized with insulin in beta cell granules and may share the proteolytic processing enzymes that generate insulin from pro-insulin.

Amylin is primarily synthesized in pancreatic beta cells and is secreted in response to nutrient stimuli such as glucose and arginine. Studies with cloned beta- cell tumor lines (Moore et al.. Biochem . Biophys . Res . Commun . , 179(1) (1991)), isolated islets (Kanatsuka et al. , FEBS Letts . , 259(1), 199-201 (1989)) and perfused rat pancreases (Ogawa et. al . , J. Clin . Invest . , 85:973-976

(1990) ) have shown that short pulses, 10 to 20 minutes, of nutrient secretagogues such as glucose and arginine, stimulate release of amylin as well as insulin. The molar amylin:insulin ratio of the secreted proteins varies between preparations from about 0.01 to 0.4, but appears not to vary much with acute stimuli in any one prepara¬ tion. However, during prolonged stimulation by elevated glucose, the amylin: insulin ratio can progressively increase (Gedulin et. al . , Biochem . Biophys . Res . Commun . , 180 (1) :782-789 (1991)) . Thus, perhaps because gene expression and rate of translation are independently controlled, amylin and insulin are not always secreted in a constant ratio.

It has been discovered that certain actions of amylin are similar to known non-metabolic actions of CGRP and calcitonin; however, the metabolic actions of amylin discovered during investigations of this newly identified protein appear to reflect its primary biologic role. At least some of these metabolic actions are mimicked by CGRP, albeit at doses which are markedly vasodilatory (see, e.g. , Leighton et al. , Nature, 335:632-635 (1988)) ; Molina et al. , Diabetes, 39:260-265 (1990)) .

The first discovered action of amylin was the reduction of insulin-stimulated incorporation of glucose into glycogen in rat skeletal muscle (Leighton et_. al . , Na ture, 335:632-635 (1988)) ; the muscle was made "insulin- resistant." Subsequent work with rat soleus muscle ex- vivo and in vitro has indicated that amylin reduces glycogen synthase activity, promotes conversion of glyco¬ gen phosphorylase from the inactive b form to the active a form, promotes net loss of glycogen (in the presence or absence of insulin) , increases glucose-6-phosphate levels, and can increase lactate output (see, e.g. , Deems et. al. , Biochem . Biophys . Res . Commun . , 181 (1) : 116-120 (1991)) ; Young et aJL, FEBS Letts, 281 (1, 2) :149-151 (1991)) . Amylin, like epinephrine, appears not to affect glucose transport per se (e.g. , Pittner et_. al . , FEBS Letts . , 365 (1) :98-100 (1995)) . Studies of amylin and insulin dose-response relations show that amylin acts as a non- competitive or functional antagonist of insulin in skele¬ tal muscle (Young et al . , Am. J. Physiol . , 263(2) :E274- E281 (1992) ) . There is no evidence that amylin interferes with insulin binding to its receptors, or the subsequent activation of insulin receptor tyrosine kinase (Follett et al . , Clinical Research, 39(1) :39A (1991)) ; Koopmans et al. , Diabetologia , 34:218-224 (1991)) . The actions of amylin on skeletal muscle resemble those of adrenaline

(epinephrine) . However, while adrenaline's actions are believed to be mediated largely by cAMP, some workers have concluded that amylin' s actions are not mediated by cAMP (see Deems et al . , Biochem . Biophys . Res . Commun . , 181(1)-

:116-120 (1991)) . Others report that amylin does activate adenyl cyclase and increases cAMP in skeletal muscle

(Pittner et al . , Biochim Biophys . Acta , 1267:75-82 (1995) ; Moore and Rink, Diabetes, 42 (Suppl 1) :257A (abstract 821)

(1993)), consistent with transduction of its effect on glycogen metabolism via cAMPdependent protein kinase phosphorylation of synthase and phosphorylase.

It is believed that amylin acts through recep- tors present in plasma membranes. It has been reported that amylin works in skeletal muscle via a receptor- mediated mechanism that promotes glycogenolysis, by activating the rate-limiting enzyme for glycogen break¬ down, phosphorylase a (Young et al. , FEBS Letts . , 281:149- 151 (1991) ) . Studies of amylin and CGRP, and the effect of selective antagonists, suggest that amylin acts via its own receptor (Beaumont et al . , Br. J. Pharmacol . , 115(5) :- 713-715 (1995) ; Wang et al . ■ FEBS Letts . , 219:195-198 (1991 b) ) , counter to the conclusion of other workers that amylin may act primarily at CGRP receptors (e.g. , Chantry et al. , Biochem . J. , 277:139-143 (1991)) ; Galeazza et al.. Peptides , 12:585-591 (1991)) ; Zhu et al. , Biochem. Biophy- s . Res . Commun . , Ill { 2 ) -. Ill- lie (1991)) . Recently, amylin receptors and their use in various methods for screening and assaying for amylin agonist and antagonist compounds were described in United States Patent No. 5,264,372, issued November 23, 1993.

While amylin has marked effects on hepatic fuel metabolism in vivo, there is no general agreement as to what amylin actions are seen in isolated hepatocytes or perfused liver. The available data do not support the idea that amylin promotes hepatic glycogenolysis, i.e. , it does not act like glucagon (e.g. , Stephens e_t al . , Diabe¬ tes, 40:395-400 (1991) ; Gomez-Foix e_t al . , Biochem J. , 276:607-610 (1991)) . It has been suggested that amylin may act on the liver to promote conversion of lactate to glycogen and to enhance the amount of glucose able to be liberated by glucagon (see Roden et al. , Diabetologia , 35:116-120 (1992)) . Thus, amylin could act as an anabolic partner to insulin in liver, in contrast to its catabolic action in muscle.

In fat cells, contrary to its adrenalin-like action in muscle, amylin has no detectable actions on insulin-stimulated glucose uptake, incorporation of glucose into triglyceride, C0₂ production (Cooper et al .. Proc . Na tl . Acad. Sci . , 85:7763-7766 (1988)) epinephrine- stimulated lipolysis, or insulin-inhibition of lipolysis (Lupien and Young, "Diabetes Nutrition and Metabolism - Clinical and Experimental," Vol. 6(1) , pages 1318 (Febru¬ ary 1993)) . Amylin thus exerts tissue-specific effects, with direct action on skeletal muscle, marked indirect (via supply of substrate) and perhaps direct effects on liver, while adipocytes appear "blind" to the presence or absence of amylin.

It has also been reported that amylin can have marked effects on secretion of insulin. In isolated islets (Ohsawa et. al . , Biochem. Biophys . Res . Commun . , 160 (2) : 961-967 (1989)) , in the perfused pancreas (Silvest- re et al . , Reg. Pept . , 31:23-31 (1991)), and in the intact rat (Young et al. , Mol . Cell . Endocrinol . , 84:R1-R5 (1992)) , some experiments indicate that amylin inhibits insulin secretion. Other workers, however, have been unable to detect effects of amylin on isolated $-cells, on isolated islets, or in the whole animal (see Broderick et al. , Biochem . Biophys . Res . Commun . , 177:932-938 (1991) and references therein) .

In preclinical experiments, a striking effect of amylin in vivo is stimulation of a rapid rise in plasma lactate, followed by a rise in plasma glucose (Young et al.. FEBS Letts . , 281 (1,2) :149-151 (1991)) . Evidence indicates that the increased lactate provides substrate for glucose production and that amylin actions can occur independent of changes in insulin or glucagon. In "glu¬ cose clamp" experiments, amylin infusions cause "insulin resistance", both by reducing peripheral glucose disposal, and by limiting insulin-mediated suppression of hepatic glucose output (e.g. , Frontoni et al. , Diabetes, 40:568- 573 (1991)) ; Koopmans et al . , Diabetologia , 34:218-224 (1991) ) .

In lightly anesthetized rats which were fasted for 18 hours to deplete their stores of hepatic glycogen, amylin injections stimulated rises in plasma lactate from about 0.5 to 1.5 mM followed by a prolonged increase in plasma glucose levels from about 6 to 11 mM. These effects were observed for both intravenous and subcutane¬ ous injections (Young et al. , FEBS Letts . , 281 (1, 2) :149- 151 (1991) ) . The effects of amylin in fed animals differ quantitatively from its effects in fasted animals. In fed rats, with presumably normal liver glycogen stores, amylin causes a more pronounced and prolonged rise in plasma lactate; however, there is only a modest rise in plasma glucose. It has been suggested that amylin promotes the "return limb" of the Cori cycle, i.e.. muscle glycogen via breakdown to lactate provides substrate for hepatic gluconeogenesis and glycogen production and probably triglyceride synthesis. Insulin drives the forward limb, i.e., uptake of glucose into muscle and production of muscle glycogen. Insulin and amylin can thus be seen as partners in regulating the "indirect" pathway of post¬ prandial hepatic glycogen repletion. "Insulin resistance" in muscle and liver may be under normal, physiologic regulation by amylin.

Amylin or amylin agonists potently inhibit gastric emptying in rats (Young et al .. Diabetologia 38 (6) :642-648 (1995)) , dogs (Brown et al. , Diabetes 43 (Suppl 1) :172A (1994)) and humans (Macdonald et al. , Diabetologia 38 (Suppl 1) :A32 (abstract 118) (1995)) . Because gastric emptying is accelerated in amylin-defi- cient type 1 diabetic BB rats (Young et al . , Diabetologia , supra; Nowak et al . , J. Lab. Clin . Med . , 123(1) :110-6 (1994) ) and in rats treated with the selective amylin antagonist, AC187 (Gedulin et al. , Diabetologia, 38(Suppl- 1) :A244 (1995) , the effect of amylin on gastric emptying appears to be physiological (operative at concentrations that normally circulate) .

Control of nutrient release from the stomach is becoming recognized as an important component of overall fuel homeostasis. In human volunteers, over a range of carbohydrate concentrations in a liquid meal, energy release from the stomach was remarkably constant at ~2 kcal/min (Brener e al . , Gastroenterology, 85(l) :76-82 (1983)), equivalent to -500 mg glucose/min. This release rate is about the same as the rate of glucose disposal that insulin-sensitive individuals can attain at their peak plasma insulin concentrations (1 nM) (Young et_. al. , Am. J. Physiol . , 254 (2 Pt 1) :E231-236 (1988)) . Thus, the rate at which carbohydrate is released from the stomach and absorbed is normally matched to the rate at which it can be metabolized.

Several feedback loops may control nutrient efflux from the stomach. Peptides that are candidates as participants in such control loops should (1) change in response to meals, and (2) potently modulate gastric emptying. Peptides known to inhibit gastric emptying are limited to amylin, cholecystokinin (CCK) , glucagon-like peptide-1 (GLP-1) , secretin and gastrin releasing pep- tide/bombesin (GRP) . GRP does not change with meals. Secretin is secreted in response to acid (but not nutri¬ ents) entering the duodenum. Of the peptides that are secreted in response to nutrients (amylin, CCK and GLP-1) , only amylin and GLP-1 are secreted in response to glucose ingestion. CCK is secreted in response to fat ingestion. of the hormones that might therefore mediate feedback control of the gastric release and subsequent absorption of glucose (amylin and GLP-1) , amylin is the more po¬ tent (Young et al . , Metabolism Clinical and Experimental 45(1) :l-3 (1996)), and it appears that amylin may be a major regulator of carbohydrate absorption, at least in rodents (Young et al . , Biochemical Society Transactions 23 (2) :325-331 (1995) ) .

Non-metabolic actions of amylin include vasodi¬ lator effects which may be mediated by interaction with CGRP vascular receptors . Reported in_. vivo tests suggest that amylin is at least about 100 to 1000 times less potent than CGRP as a vasodilator (Brain et al . , Eur. J. Pharmacol . , 183:2221 (1990) ; Wang et al . , FEBS Letts . , 291:195-198 (1991)) . The effect of amylin on regional hemodynamic actions, including renal blood flow, in conscious rats was recently reported (Gardiner et al.. Diabetes, 40:948-951 (1991)) . The authors noted that infusion of rat amylin was associated with greater renal vasodilation and less mesenteric vasoconstriction than is seen with infusion of human "-CGRP. They concluded that, by promoting renal hyperemia to a greater extent than did "-CGRP, rat amylin could cause less marked stimulation of the renin-angiotensin system, and thus, less secondary angiotensin II-mediated vasoconstriction. It was also noted, however, that during co-infusion of human "-^8_37CGRP and rat amylin renal and mesenteric vasoconstrictions were unmasked, presumably due to unopposed vasoconstrictor effects of angiotensin II, and that this finding is similar to that seen during co-infusion of human "-CGRP and human "-⁸-³⁷CGRP (id. at 951) . Injected into the brain, or administered periph¬ erally, amylin has been reported to suppress food intake e.g. , Chance et al■ , Brain Res . , 539:352-354 (1991)), an action shared with CGRP and calcitonin. The effective concentrations at the cells that mediate this action are not known. Amylin has also been reported to have effects both on isolated osteoclasts where it caused cell quies- cence, and in vivo where it was reported to lower plasma calcium by up to 20% in rats, in rabbits, and in humans with Paget's disease (see, e.g.. Zaidi e al . , Trends in Endocrinol . and Metab. , 4:255-259 (1993) . From the avail- able data, amylin seems to be 10 to 30 times less potent than human calcitonin for these actions. Interestingly, it was reported that amylin appeared to increase osteo- clast cAMP production but not to increase cytosolic Ca²⁺, while calcitonin does both (Ala et al. , Biochem. Biophys . Res . Commun . , 179 (1) :134-139 (1991)) . It was suggested, though not established, that calcitonin may act via two receptor types and that amylin may interact with one of these .

The infusion of amylin receptor antagonists may be used to alter glucoregulation. ^8~3 CGRP is a demonstrated amylin blocker in vitro and in vivo (Wang et al . , Biochem . Biophys . Res . Commun . , 181 (3) :1288-1293 (1991)) , and was found to alter glucoregulation following an arginine infusion in fed rats (Young et al .. Mol . Cell . Endocrinol - ., 84:R1-R5 (1992)) . The initial increase in glucose concentration is attributed to arginine-stimulated gluca- gon secretion from islet alpha cells; the subsequent restoration of basal glucose is attributed to insulin action along with changes in other glucoregulatory hor- mones. When the action of amylin is blocked by preinfusi- on of ^8"37hCGRP, the initial glucose increase is not significantly different, but there is a subsequent fall in glucose concentration to well below the basal level, which is restored only after some 80 minutes. Thus, glucoregul- ation following this challenge with an islet secretagogue was reportedly altered by infusion of an amylin receptor antagonist. Additionally, insulin concentrations were measured at half-hour intervals and it was found that insulin concentration 30 minutes following the arginine infusion was almost twice as high in animals infused with an amylin receptor antagonist as in the normal controls . ^8~37CGRP is also an effective CGRP antagonist. However, very similar results were seen with another amylin antago¬ nist, AC66, which is selective for amylin receptors compared with CGRP receptors (Young et al . , Mol . Cell . Endocrinol . , 84:R1-R5 (1992)) . These results are said to support the conclusion that blockade of amylin action can exert important therapeutic benefits in Type 2 diabetes .

It has also been discovered that amylin markedly increases plasma renin activity in intact rats when given subcutaneously in a manner that avoids any disturbance of blood pressure. This latter point is important because lowered blood pressure is a strong stimulus to renin release. Amylin antagonists, such as amylin receptor antagonists, including those selective for amylin recep- tors compared to CGRP and/or calcitonin receptors, can be used to block the amylin-evoked rise of plasma renin activity. These findings support the determination that amylin antagonists will reduce plasma renin activity with consequent therapeutic benefit in hypertension and cardiac failure and other disorders associated with elevated, inappropriate or undesired renin activity. Moreover, the additional ability of amylin antagonists to favorably modulate insulin resistance and other common metabolic disorders frequently associated with hypertension and cardiac disease provides a particularly desirable thera¬ peutic profile. The use of amylin antagonists to treat renin-related disorders is described in United States Patent No. 5,376,638, issued December 27, 1994.

In normal humans, fasting amylin levels from 1 to lOpM and post-prandial or post-glucose levels of 5 to 20pM have been reported (e.g. , Hartter et_. al . , Diabetolog¬ ia , 34:52-54 (1991) ; Sanke et al. , Diabetologia, 34:129- 132 (1991) ; Koda et al. , The Lancet , 339:1179-1180 (1992)- ) . In obese, insulin-resistant individuals, post-food amylin levels can go higher, reaching up to about 50pM. For comparison, the values for fasting and post-prandial insulin are 20 to 50pM, and 100 to 300 pM respectively in healthy people, with perhaps 3-to 4-fold higher levels in insulin-resistant people. In Type 1 diabetes, where beta cells are destroyed, amylin levels are at or below the levels of detection and do not rise in response to glucose (Koda e al. , The Lancet, 339:1179-1180 (1992)) . In normal mice and rats, basal amylin levels have been reported from 30 to 100 pM, while values up to 600 pM have been measured in certain insulin-resistant, diabetic strains of rodents (e.g. , Huang et al. , Hypertension, 19:1-101-1-109 (1991) ; Gill et al.. Life Sciences, 48:703- 710 (1991) ) .

Patients with Type 1 diabetes, in addition to a lack of insulin, are reported to have marked amylin defi- ciency. As noted above, data show that amylin expression and secretion by pancreatic beta cells is absent or well below normal in Type 1 diabetes. In several animal models of Type 1 diabetes, amylin secretion and gene expression are depressed (Cooper et. al . , Diabetes , 497-500 (1991) ; Ogawa et al . , J. Clin . Invest . , 85:973-976 (1990)) . Measurements of plasma amylin in Type 1 diabetic patients show that amylin is deficient in these patients after an overnight fast, and that a glucose load does not elicit any increase in amylin levels (Koda et. al. , The Lancet, 339:1179-1180 (1992)) .

Transσenic Mammals

Various mammals have been produced which have altered levels of expression of certain genes. One class of these mammals are the so-called transgenic mammals. These mammals have a novel gene or genes introduced into their genome. Another class of these mammals is the so- called "knockout" mammals, wherein expression of an endogenous gene has been suppressed thorough genetic manipulation. A variety of transgenic mammals have been developed. For example, United States Patent No. 4,736,8- 66 issued April 12, 1988 describes a mouse containing a transgene encoding an oncogene . United States Patent No. 5,175,384, issued December 29, 1992, describes a transge¬ nic mouse deficient in mature T cells. United States Patent No. 5,175,383, issued December 29 ,1992, describes a mouse with a transgene encoding a gene in the int-2-FGF family. This gene promotes benign prostatic hyperplasia. United States Patent No. 5,175,355, issued December 29, 1992, describes a transgenic mouse with enhanced resis- tance to certain viruses. United States Patent No. 5,368,854, issued November 29, 1994, discusses raising a transgenic mouse which lacks the IL-10 gene. United States Patent No. 5,387,742, issued February 7, 1995, describes a transgenic mouse which reportedly displays the amyloid-forming pathology of Alzheimer's disease. United States Patent No. 5,489,743, issued February 6, 1996, describes a transgenic mouse which is said to exhibit decreased platelet counts and/or megakaryocyte leukemia. United States Patent No. 5,489,742, issued February 6, 1996, describes a transgenic rat which has the human HLA- B27 gene inserted in to its genome. United States Patent No. 5,491,283, issued February 13, 1996 describes a transgenic mouse which contains a transgene comprising a BCR/ABL gene fusion and which develop leukemia. Preparation of a knockout mammal requires first introducing a nucleic acid construct which will be used to suppress expression of a particular gene into an undiffer- entiated cell type termed an embryonic stem cell. This cell is then injected into a mammalian embryo, where it hopefully will be integrated into the developing embryo. The embryo is then implanted into a foster mother for the duration of the gestation.

Pfeffer et al. , Cell, 73:457-467 (1993) report mice in which the gene encoding the tumor necrosis factor receptor p55 has been suppressed. The mice showed a decreased response to tumor necrosis factor signaling. Fung-Leung et al . , Cell, 65:443-449 (1991) report knockout mice lacking expression of the gene encoding CD8. Kishira et al. , Cell, 74:143-56 (1993) report the generation of a mouse with a mutation in exon 6 on CD45. This mouse reportedly does not express CD45. A Fmrl knockout mouse has been proposed by the Dutch-Belgian Fragile X Consor¬ tium as a model to study fragile X mental retardation (Cell, 78:23-33 (1994)) . Di Simone et al_^, Transplanta¬ tion, 61:13-19 (1996) , report a alpha-1, 3-galactosyltrans- ferase knockout mouse and suggest that it may have impli¬ cations for xenotransplantation. Thomas, Am. J. Cell. Mol . Biol. , 12:461-463 (1995) discusses knockout mice and the phenotypes resulting from various mutations.

Animal Models

Several animal models exist for studying diabe¬ tes. For example, treatment of rats with streptozotocin (STZ) results in destruction of the beta cells of the pancreas and provides an animal model for the study of Type 1 diabetes mellitus. In these models, both amylin and insulin secretion and gene expression are depressed or absent (Cooper et al ■ , Diabetes, 497-500 (1991) ; Ogawa et al. , J. Clin . Invest . , 85:973976 (1990)) .

Amylin Activity Screening Methods

Current methods for screening compounds for amylin activity include the receptor binding assay which is described in United States Patent No. 5,264,372, issued November 23, 1993. Assays of amylin agonist or antago- nist activity may be performed using the soleus muscle assay previously described (Leighton, B. and Cooper, Nature, 335:632-635 (1988) ; Cooper, et al . , Proc. Na tl . Acad . Sci . USA 85:7763-7766 (1988)) . In vivo methods of determining amylin agonist or antagonist activity include methods for measuring the rate of gastric emptying, which are disclosed in, for example, Young et al . , Diabetologia. 38 (6) :642-648 (1995) , or by measuring lactate and/or glucose levels following compound administration adminis¬ tration, as described, for example, in International Patent Application No. WO 92/11863, published on July 23, 1992.

In view if the lack of an in vivo model for amylin deficiency alone, there is a need for in vivo systems studying amylin deficiency and amylin action in the presence of endogenous insulin. An animal model for amylin deficiency alone is also useful for screening and evaluating compounds useful in the treatment of amylin- related disorders, including diabetes. Accordingly, it is an object of this invention to provide mammals in which a gene involved in the production of amylin has been sup- pressed.

SUMMARY OF THE INVENTION

In one aspect, the present invention provides an animal, preferably a mouse, and its progeny having a suppressed level of expression of the gene encoding amylin. The gene may be suppressed by inserting into the relevant portion of the genome of the mouse or other animal a nucleic acid sequence comprising, for example, a portion of an exon of the amylin-encoding sequence of the animal linked to a marker sequence. The sequence into which the nucleic acid sequence is inserted may be any sequence which allows expression of amylin to be blocked, for example, a sequence which codes for preproamylin, proamylin or amylin. The marker sequence may be an antibiotic resistance gene, for example, a neomycin resistance gene.

In another embodiment, the invention provides a mouse or other animal and its progeny wherein expression of the gene encoding amylin is suppressed. The invention further provides mice and other animals with a decreased level of amylin expression. 16

In yet another aspect, the invention provides a method of screening a compound for amylin activity com¬ prising administering the compound to a mouse or other animal with a suppressed level of amylin expression, and assaying the mouse or other animal for amylin activity.

BRIEF DESCRIPTION OF THE DRAWINGS

FIGURE 1 depicts a mouse amylin genomic sequence (LAMBDA M-amylin 14) isolated from a mouse Balb/c lambda FIXII library using a human amylin cDNA probe in low stringency hybridization.

FIGURE 2 depicts a 3.46 kbp Hindlll-Xbal frag¬ ment of a subclone from LAMBDA M-amylin-14 (pHX) .

FIGURE 3 depicts the sequence resulting from removal of the amylin encoding sequence from pHX (by deleting an 890-base pair sequence by restriction diges¬ tion with SphI and BamHI) .

FIGURE 4 depicts the knockout construct used to suppress expression of amylin. FIGURE 5 depicts plasma glucose levels (Fig. 5A) and plasma insulin levels (Fig. 5B) in male and female amylin knockout mice (solid circles) and wild type mice (open circles) . Values are plus or minus the standard error of the mean. *** P < 0.001, ** P < 0.01, * P < 0.05, two-tailed unpaired t-test.

FIGURE 6 depicts plasma glucose levels (Fig. 6A) and plasma insulin levels (Fig. 6B) in male and female amylin knockout mice (solid circles) and wild type mice

(open circles) , during OGTT. Values are plus or minus the standard error of the mean. *** P < 0.001, ** P < 0.01,

* P < 0.05, two-tailed unpaired t-test.

FIGURE 7 depicts plasma glucose levels (Fig. 7A) and plasma insulin levels (Fig. 7B) in male and female amylin knockout mice (solid circles) and wild type mice (open circles) , during IVGTT. Values are plus or minus the standard error of the mean. *** P < 0.001, ** p < 0.01, * P < 0.05, two-tailed unpaired t-test.

FIGURE 8 depicts approximated daily food intake (Fig. 8A) and weight gain Fig. 8B) in male and female amylin knockout mice (solid circles) and wild type mice (open circles) . Values are plus or minus the standard error of the mean. *** P < 0.001, ** P < 0.01, * P < 0.05, two-tailed unpaired t-test.

FIGURE 9 depicts the duration of paw licking in the first (0-5 minutes) and second phase (10-30 minutes) following subcutaneous injection of formalin in amylin knockout (mutant) mice and wild type mice. * P = 0.027,

Mann-Whitney U-test.

DETAILED DESCRIPTION OF THE INVENTION

The term "knockout" refers to a partial or complete suppression of the expression of at least a portion of a protein encoded by an endogenous DNA sequence in a cell. The present invention provides knockout mammals in which the expression of the protein amylin is suppressed.

A knockout construct nucleic acid sequence may comprise a full or partial sequence of one or more exons and/or introns of the gene to be suppressed and/or a full or partial promotor sequence of the gene to be suppressed. As used herein, the term "knockout construct" refers to a nucleic acid sequence that is designed to decrease or suppress expression of a protein encoded by endogenous DNA sequences in a cell, and is typically comprised of (1) DNA from some portion of the amylin gene; and (2) a marker sequence used to detect the presence of the knockout nucleic acid sequence in the cell. The term amylin gene refers to any DNA sequence relating to the production of amylin, including but not limited to the prepro-amylin sequence, the pro-amylin sequence, and the amylin sequence currently found on chromosome 12. The term "marker se- quence" refers to a nucleic acid sequence that is (1) used as a part of a knockout nucleic acid construct to disrupt the expression of the amylin gene; and (2) used as a means to identify those cells that have incorporated the knock- out construct into the genome. The marker sequence may be any sequence which serves these purposes, although typi¬ cally it will be a sequence encoding a protein that confers a detectable trait on the cell, such as an antibi¬ otic resistance gene or an assayable enzyme not typically found in the cell. If the marker sequence encodes a protein, the marker sequence will also typically contain a promotor which regulates its expression.

The knockout construct is inserted into a cell, and integrates with the genomic DNA of the cell in such a position so as to prevent or interrupt transcription of the native DNA sequence. Such insertion usually occurs by homologous recombination. Thus, regions of the knockout construct that are homologous to endogenous DNA sequences hybridize to each other when the knockout construct is inserted into the cell and recombine so that the knockout construct is incorporated into the corresponding position of the endogenous DNA. Typically, the knockout construct is inserted into an embryonic stem cell and is integrated into the embryonic stem cell genomic DNA, usually by the process of homologous recombination. This embryonic stem cell is then injected into, and integrates with, the developing embryo.

The term "amylin" refers to the 37-amino acid protein hormone having a disulphide bridge between the cysteine residues at positions two and seven and a C- terminal NH₂ group. See United States Patent No. 5,367,05- 2, issued November 22, 1994.

The term "progeny" refers to any and all future generations derived and descending from a particular mammal, i.e., a mammal containing a knockout construct inserted in to its genomic DNA. Thus, progeny of any successive generation are included herein such that the progeny, the FI, F2 and F3 generations and so on indefi¬ nitely are included in this definition.

The term "amylin activity" refers to a biologi- cal activity of amylin, for example, activity in inhibit¬ ing insulin-stimulated glycogen synthesis in isolated soleus muscle, or delaying gastric emptying in vivo .

Usually, the DNA to be used in the knockout construct will be a portion or all of one or more amylin exon and/or intron regions, with or without a promotor region. Generally, the DNA will be at least about 1 kil- obase pair (kbp) in length and preferably 3-4 kbp or more in length, more preferably 5-14 kbp or more in length and, specifically, is as long or as short as needed to provide for a sufficient complementary sequence for hybridization when the knockout construct is introduced into the genomic DNA of the embryonic stem cell. As shown in Example 1, the DNA to be used in preparing the knockout construct contained exon three of the amylin gene. The DNA sequence to be used to knock out the amylin gene can be obtained using various art known methods such as those described by Sambrook e al. , Mol ecular Cloning: A Laboratory Manual , Cold Spring Harbor Laboratory PRess, Cold Spring Harbor, NY (1989) . Such methods include, for example, screening a genomic library with a cDNA probe encoding at least a portion of the amylin gene to obtain at least a portion of the genomic sequence for amylin. Example 1 describes the use of a human amylin cDNA probe corresponding to a 300 bp BamHI fragment of a cDNA clone (described Mosselman, et al . , FEBS Lett . 247: 154-158 (1989)) to screen a mouse Balb/c lambda FIXII library in order to obtain a DNA fragment containing a desired portion of the amylin gene. If a promotor is to be used in the knockout construct, synthetic DNA probes can be designed for screening a genomic library containing the amylin promotor sequence. Another method for obtain- ing the DNA to be used in the knockout construct is to manufacture the DNA sequence synthetically, using a DNA synthesizer.

The DNA sequence encoding the knockout construct must be generated in sufficient quantity for genetic manipulation and insertion into embryonic stem cells. Amplification may be conducted by known methods, including (1) placing the sequence into a suitable vector and trans¬ forming bacterial or other cells which can rapidly amplify the vector; (2) by polymerase chain reaction (PCR) ampli¬ fication; or (3) by synthesis with a DNA synthesizer.

The DNA sequence to be used in producing the knockout construct is digested with a particular restric¬ tion enzyme selected to cut at a location(s) such that a new DNA sequence encoding a marker sequence can be insert¬ ed in the proper position within this DNA sequence. The proper position for marker sequence insertion is that which will serve to prevent expression of the native amylin gene. This position will depend on various fac- tors, such as the restriction sites in the sequence to be cut, and whether an exon sequence or a promotor sequence, or both ,are to be interrupted (i.e. , the precise location of insertion necessary to inhibit promotor function or to inhibit synthesis of the native exon) . Preferably, the enzyme selected for cutting the DNA will generate a longer arm and a shorter arm, where the shorter arm is about 300 base pairs (bp) . In some cases, it will be desirable to actually remove a portion or even all of one or more exons of the amylin gene so as to keep the length of the knockout construct comparable to the original genomic sequence when the marker sequence is inserted in the marker construct. In these cases, the genomic amylin DNA is cut with appropriate restriction endonucleases such that a fragment of the proper size can be removed. As shown in Example 1, the amylin-encoding sequence was deleted from the exon 3-containing DNA sequence by re- striction enzyme digestion. The selected marker sequence was then inserted in its place.

The marker sequence can be any nucleic acid sequence which is detectable and/or assayable, however typically it is an antibiotic resistance gene or other sequence whose expression or presence in the genome can easily be detected. The marker sequence is usually opera¬ bly linked to its own promoter or to another strong promo¬ tor from any source which will be active or which can easily be activated in the cell into which it is inserted. The marker sequence need not, however, have its own promo¬ tor attached as it may be transcribed using the promotor of the amylin gene. In addition, the marker sequence will normally have a polyA sequence attached to the 3' end, which serves to terminate transcription. Preferred marker genes are any antibiotic resistance genes such as neo

(neo ycin resistance gene) or genes coding for an enzyme having detectable activity such as beta-gal (beta-galacto- sidase) . After the genomic DNA sequence has been digested with the appropriate restriction enzymes, the marker sequence is ligated in to the genomic DNA sequence using methods known to those skilled in the art and described in Sambrook et al . supra. The ends of the DNA fragments to be ligated must be compatible; this is achieved by either cutting all fragments with enzymes which generate compati¬ ble ends, or by blunting the ends prior to ligation (for example by the use of Klenow fragment to fill in sticky ends) . As shown in Example 1, a knockout construct was constructed to include a phosphoglycerate kinase 1 promot¬ er-driven neomycin resistance gene cassette as the marker sequence.

The ligated knockout construct may then be inserted directly into embryonic stem cells, or it may be first be placed into a suitable vector for amplification prior to insertion. Preferred suitable vectors are those which are rapidly amplified in bacterial cells such as the pBluescript II SK vector (Stratagene, San Diego, CA) or pGEM7 (Promega Corp., Madison, WI) .

This invention contemplates the production of knockout mammals, such as mammals from any species of rodent, including without limitation, rabbits, rats, hamsters, and mice. Preferred rodents include members of the Muridae family, which includes rats and mice. Gener¬ ally, the embryonic stem cells used to produce the knock- out mammal will be of the same species as the knockout mammal to be generated. Thus, for example, mouse embryon¬ ic stem cells will usually be used for generation of knockout mice. The use of embryonic stem cells from mice to generate amylin knockout mice is described in Example 2.

Embryonic stem cells are typically selected for their ability to integrate into and become part of the germ line of a developing embryo so as to create germ line transmission of the knockout construct. Thus, any embry- onic stem cell line which is believed to have this capa¬ bility is suitable for use herein. One mouse strain which is typically used for the production of embryonic stem cells is the 129J strain. A preferred embryonic stem cell line is murine cell line D3 (American Type Culture Collec- tion catalog no. CRL 1934) . The mouse 129SV embryonic stem cell line E14 (Hooper et al .. Nature, 326:292-295 (1987) ) was used in the experiments described in Example 2.

The cells are cultured and prepared for DNA insertion using methods known in the art, such as those set forth by Robertson in: Teratocarcinomas and Embryonic Stem Cells: A Practical Approach, E.J. Robertson ed. , IRL PRess, Washington, D.C. (1987) ; and by Hogan et al.. Manipulating the Mouse Embryo: A Laboratory Manual. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY (1986) . Insertion of the knockout construct into the embryonic stem cells can be accomplished using a variety of methods known in the art including, for example, electroporation, microinjection, and calcium phosphate treatment (see Lovell-Badge, in Robertson, ed. , supra) . A preferred method of insertion is electroporation, which was used in the experiments described in Example 2.

As described in Example 2, each knockout con¬ struct DNA to be inserted into the cell is first lineariz¬ ed if the knockout construct has been inserted into a vector, which is accomplished by digesting the DNA with a suitable restriction endonuclease selected to cut only within the vector sequence and not within the knockout construct sequence. For insertion of the DNA sequence, the knockout construct DNA is added to the embryonic stem cells under appropriate conditions for the insertion method chosen. If the cells are to be electroporated (see Example 2) , the embryonic stem cells and knockout con¬ struct DNA are exposed to an electric pulse using an electroporation machine and following the manufacturer's guidelines for use. After electroporation, the cells are allowed to recover under suitable incubation conditions. The cells are then screened for the presence of the knockout construct.

Screening can be done using a variety of meth- ods. If the marker sequence is an antibiotic resistance gene, the cells are cultured in the presence of an other¬ wise lethal concentration of antibiotic. Those cells which survive have presumably integrated the knockout construct. In order to identify cells to be used to prepare the amylin knockout mice described herein, recom- binant embryonic stem cells were evaluated for resistance to neomycin. If the marker sequence is other than an antibiotic resistance sequence, a Southern blot of the embryonic stem cell genomic DNA can be probed with a sequence of DNA designed to hybridize only to the marker sequence. If the marker sequence is a gene which encodes an enzyme whose activity can be detected (for example, beta-galactosidase) , the enzyme substrate can be added to the cells under suitable conditions, and the enzymatic activity can be analyzed. The knockout construct may be integrated into several locations in the embryonic stem cell genome, and may integrate into a different location in each cell's genome, due to the occurrence of random insertion events. The desired location of the insertion is in a complementa- ry position to the DNA sequence to be knocked out. Typically, less than about 1-5 percent of the embryonic stem cells which take up the knockout construct will actually integrate the knockout construct in the desired location. To identify those cells with proper integration of the knockout construct, the DNA can be extracted from the cells using standard methods such as those described by Sambrook et al. , supra. As referenced in Example 2, the DNA can then be probed on a Southern blot with a probe or probes designed to hybridize in a specific pattern to genomic DNA digested with particular restriction en¬ zyme (s) . Alternatively or additionally, the genomic DNA can be amplified by PCR with probes specifically designed to amplify DNA fragments of a particular size and sequence (i.e. , only those cells which contain the knockout con- struct in the proper position will generate DNA fragments of the proper size) .

After suitable embryonic stem cells containing the knockout construct in the proper location have been identified, the cells are inserted into an embryo. Inser- tion may be accomplished in a variety of ways, however, a preferred method is by microinjection, as set forth in Example 2. For microinjec-tion, about 10-30 cells are collected into a micropipet and injected into embryos in the proper stage of development to integrate the embryonic stem cell into the developing embryo. The suitable stage of development for the embryo is generally species depen- dent; for mice it is about 3.5 days. The embryos may then be obtained by perfusing the uterus of pregnant females. Suitable methods for accomplishing this are known in the art and are set forth by Bradley in Robertson, ed. , supra. While any embryo of the proper age and stage of develop¬ ment is suitable for use, preferred embryos are male and have genes coding for a coat color which is different from the coat color encoded by the embryonic stem cell genes. In this way, the offspring can be screened easily for the presence of the knockout construct by looking for mosaic coat color (indicating that the embryonic stem cell was incorporated into the developing embryo) . See Example 2. Thus, for example, if the embryonic stem cell line carries the genes for white fur, the embryo selected will carry genes for black or brown fur.

After the embryonic stem cell has been intro¬ duced into the embryo, the embryo is implanted into the uterus of a pseudopregnant foster mother animal. While any foster mother animal may be used, they are typically selected for their ability to breed and reproduce well and for their ability to care for their young. Such foster mother animals are typically prepared by mating with vasectomized males of the same species. The stage of the pseudopregnant foster mother animal is important for successful impartation, and it is species dependent. For mice, this state is about 2-3 days.

As noted above and in Example 2, offspring which are born to the foster mother animal may be screened initially for mosaic coat color if the coat color selec- tion strategy has been employed. In addition or in the alternative, DNA from tail tissue of the offspring may be screened for the presence of the knockout construct using Southern blots and/or PCR. Offspring which appear to be mosaics are then crossed with each other if they are believed to carry the knockout construct in their germ line to generate homozygous knockout animals. If it is unclear whether the offspring will have germ line trans¬ mission, they can be crossed with a parental or other strain and the offspring screened for heterozygosity. The heterozygotes are identified by Southern blots and/or PCR amplification of the DNA. The heterozygotes can then be crossed with each other to generate homozygous knockout offspring. See Example 2.

Homozygotes may be identified by Southern blotting of equivalent amounts of genomic DNA from animals which are known to be the product of the cross, as well as animals which are known heterozygotes and wild type ani¬ mals .

Other means of identifying and characterizing the knockout offspring are available. For example, Northern blots can be used to probe the mRNA for the presence or absence of transcripts encoding either the amylin gene, the marker sequence or both. In addition, Western blots can be used to assess the level of expres¬ sion of the gene knocked out in various tissues of these offspring by probing the Western blot with an antibody against amylin, or an antibody against the marker sequence product. Finally, in situ analysis (such as fixing the cells and labelling with antibody) and/or fluorescence activated cell sorting (FACS) analysis of various cells from the offspring can be conducted using suitable anti¬ bodies to look for the presence or absence of amylin.

We have discovered that mice prepared and bred as described herein lack a functional amylin gene, as con¬ firmed by studies showing the absence of immunoreactive amylin in the islets. The mice do, however, retain normal insulin levels. It was also discovered that the amylin knockout mice develop and reproduce normally. Further¬ more, as described in Example 3, weight gain, food intake, fasting plasma levels of glucose, and basal insulin levels did not differ between the wild type and amylin knockout mice. Importantly, it was also surprisingly found that, at thirty minutes following an intravenous glucose chal¬ lenge, knockout mice had an accelerated elimination of glucose, which was further accentuated during the chal¬ lenge. See Example 3. Similarly, following a peroral glucose challenge (150 mg per animal) , glucose elimination was enhanced in knockout mice. No difference was seen in plasma insulin levels between amylin knockout mice and wild type mice in these experiments.

In further experiments, as described in Example 4, it was demonstrated that amylin knockout mice are normoglycemic in the fed and fasted state. However, as described in Example 5, the mice clear glucose more effi¬ ciently in glucose tolerance tests. Additionally, as noted in Example 5, male amylin knockout mice show abnor- mal insulin levels during glucose tolerance tests, or after a 24-hour fast. As described in Example 6, male amylin knockout mice develop obesity on a normal chow diet. As indicated in Example 4, aged amylin knockout mice, including obese males, have 25-50% reduced fed plasma insulin levels compared to controls. Additionally, as described in Example 7, amylin knockout mice react less to pain.

The amylin knockout mammals may be used to identify or evaluate amylin agonist molecules. Screening for useful compounds may include the steps of administer¬ ing a candidate compound or compounds over a range of doses to the mammal, and assaying at various time points for the effects of the compound on the amylin function being evaluated. Such assays may include, for example, looking for increased or decreased levels of amylin activities, such as hyperglycemic or hyperlactemic activi¬ ty or activity in the regulation of gastric emptying, or looking for increased or decreased levels of chemical messengers, such as cAMP. In addition, the mammals of the present invention are useful for evaluating the develop¬ ment of diabetes, the effects of loss of endogenous amylin protein activity, and for studying the effects of particu¬ lar gene mutations, including the effects of the loss of amylin gene activity.

To assist in understanding the present inven- tion, the following Examples are included which describe the results of a series of experiments. The experiments relating to this invention should not, of course, be construed as specifically limiting the invention and such variations of the invention, now known or later developed, which would be within the purview of one skilled in the art are considered to fall within the scope of the inven¬ tion as described herein and hereinafter claimed.

EXAMPLE 1 GENERATION OF CONSTRUCT

A mouse amylin genomic sequence (LAMBDA M- amylin-14) shown in FIGURE 1 was isolated from a mouse Balb/c lambda FIXII library (Stratagene, San Diego, CA) using a human amylin cDNA probe corresponding to a 300 bp

BamHI fragment of a cDNA clone (described Mosselman, et al., FEBS Lett . 247: 154-158 (1989) Jin low stringency hybridization. A 10.5 kilobase pair (kbp) continuous sequence was mapped with restriction enzyme digestion and the position of exon 3 of the amylin gene was located by Southern hybridization and DNA sequencing.

A 3.46 kbp Hindlll-Xbal fragment shown in FIGURE 2 was subcloned in plasmid pBluescript SK- (Stratagene, San Diego, CA) . This subclone (pHX) was further charac- terized and was shown to contain the amylin encoding sequence. This sequence was removed from pHX by deleting an 890-base pair sequence by restriction digestion with SphI and BamHI (see FIGURE 3) .

The deleted sequence was replaced by a phospho- glycerate kinase 1 promotor (PGK) -driven neomycin resis¬ tance gene cassette (PGK-neo) (Dr. J.K. Heath, Oxford) excised from its pBluescript SK-vector by Hindlll-Xhol, in a ligation reaction following blunt ending of the respec¬ tive DNA ends by the Klenow fragment of DNA polymerase I. The resulting DNA construct, pHX) exon3-PGKneo (FIGURE 3) was fused with a 5.1 kilobase pair Xbal-Sall fragment pXXS from LAMBDA M-amylin-14 (see FIGURE 1) to generate the targeting construct p)am (FIGURE 4) .

EXAMPLE 2 DERIVATION OF KNOCKOUT MICE

The p)am targeting construct was excised from the Bluescript vector with Xhol digestion (FIGURE 4) , purified by gel filtration and 50 Fg of the construct was electroporated into the mouse 129SV embryonic stem cell line E14 (Hooper et al. , Nature, 326:292-295 (1987)) using BioRad Gene Pulser equipment set at 260V, 500 FF. G418- resistant embryonic stem cell clones were selected, isolated, frozen and processed for Southern blot analysis as described in Leveen et al. , Genes & Development, 8:1875-1887 (1994) . A 1.5 kilobase pair EcoRI-Hindlll genomic fragment (from pEII, see FIGURE 1) , flanking the targeting construct by the 5' end, was used to identify clones which had undergone homologous recombination at the amylin gene locus. A recombinant clone (clone 60) , displaying a 3.5 kbp EcoRI fragment in addition to the wild type 12 kbp EcoRI fragment was thawed, cultured and injected into C57B6J mouse strain blastocysts as described in Leveen et al. , supra. Male chimeras were identified by virtue of their fur coat color and high degree chimeras were bred with C57B6J females. Mutant heterozygote offspring were identified by Southern blot analysis of DNA from tail biopsies. Heterozygous animals were crossed and mutant homozygotes (amylin gene knockout mice) were derived in agreement with Mendelian allele transmission. The absence of amylin-encoding genomic sequences in the amylin gene knockout mice was confirmed by Southern blot analysis using an amylin gene exon 3-specific probe. The results showed a reduced signal in the +/- compared with the +/+ DNA and absence of signal in the -/- DNA.

EXAMPLE 3 GLUCOSE CHALLENGE STUDIES

Amylin knockout mice obtained by the methods described in Examples 1 and 2 and control (wild type) mice were subjected to intravenous and oral glucose challenge tests. In the intravenous challenge test, 1 g/kg glucose was administered intravenously and the plasma glucose measured over time. In the peripheral glucose challenge test, 150 mg/animal glucose was administered orally and the plasma glucose measured over time.

Thirty minutes after intravenous glucose chal¬ lenge, knockout mice exhibited accelerated elimination of blood glucose (11.2 +/- 0.8 nM for knockout mice versus 14.8 +/- 0.7 mM (p = 0.003;n = 17)) . Similarly, 120 minutes following oral glucose challenge glucose elimina¬ tion was enhanced in knockout mice (22.0 +/- 2.4 versus 30.8 +/- 3.2 mM (p < 0.05; n = 10)) .

No difference was seen in plasma insulin levels between knockout mice and wild type mice (peak levels in intravenous glucose challenge: 1943 +/- 292 versus 1663 +/- 153 pM (p = 0.49;n = 17)) .

Nor was any difference noted between knockout and wild type animals with respect to weight gain, food intake, fasting levels of glucose (5.5 +/- 0.4 versus 5.4 +/- 0.7 nM (n = 10)) , and basal insulin levels (241 +/- 32 versus 309 +/- 66 pM (n = 10) ) . EXAMPLE 4 MEASUREMENT OF GLUCOSE AND INSULIN LEVELS

Amylin knockout mice obtained by the methods described in Examples 1 and 2 and control (wild type) mice were fed a standard chow diet (R3, LABFOR, 5% fat) and monitored from 6 to 42 weeks of age. No difference was seen in plasma glucose levels between amylin knockout mice and wild type mice (Fig. 5A) . Amylin knockout mice, 24 weeks or older, displayed a 25-50% reduction in plasma insulin levels compared to wildtype control animals (Fig. 5B) .

EXAMPLE 5

GLUCOSE CHALLENGE STUDIES

Amylin knockout mice obtained by the methods described in Examples 1 and 2, along with wild type mice, were subjected to oral glucose tolerance tests (OGTTs) and intravenous glucose tolerance tests (IVGTTs) . Amylin knockout mice were found to clear glucose more rapidly than wild type controls in both OGTT and IVGTT (Figs. 6 and 7) . In the OGTT, groups of age- and sex-matched 8-

12-week-old mice were fasted for 24 hours, anaesthetized and given 150 mg/animal glucose orally. In the IVGTT, groups of age- and sex-matched 8-12-week-old mice were anaesthetized and given 1 g/kg glucose into the tail vein without prior fasting, and the plasma glucose measured over time. Glucose and insulin concentrations in plasma were measured over time using the glucose oxidase tech¬ nique and a radioimmunoassay, respectively.

Male amylin knockout mice, but not females, showed an increase in plasma insulin in response to glucose (Figs. 6 and 7) . This is consistent with data showing that after a 24-hour fast, plasma insulin levels were more than 3-fold increased in male amylin knockout mice (but not females) compared to controls (data not shown) .

EXAMPLE 6 FOOD INTAKE AND BODY MASS STUDIES

Amylin knockout mice obtained by the methods described in Examples 1 and 2 and control (wild type) mice were fed a standard chow diet (R3, LABFOR, 5% fat) and monitored from 6 to 42 weeks of age. Food intake was not quantitatively altered in amylin knockout mice compared to wild type mice (Fig. 8A) . Nevertheless, male amylin knockout mice showed a 20% increase in body mass compared to controls from 18-42 weeks of age (Fig. 8B) . Addition¬ ally, in other experiments (data not shown) , epididy al fat pads from 20-week-old amylin knockout males were enlarged showing hypertrophic adipocytes with increased triglyceride content relative to adipocytes from wild type males .

EXAMPLE 7 NOCICEPTION STUDIES

Amylin knockout mice obtained by the methods described in Examples 1 and 2, along with wild type mice, were subjected to the formalin test, which produces a biphasic nociceptive response in rodents. See Tjolsen et al., Pain 51:5-17 (1992) . Eight-week-old males were injected subcutaneously with 20 μl of a 2% phosphate buffered paraformaldehyde solution (pH 7.2) under the dorsal surface of the right hind paw. The total time of licking during the early phase (0-5 minutes after injec- tion ) and the late phase (10-30 minutes after injection ) was determined. Amylin knockout mice displayed a 30% reduction in the late phase response, but the early phase response was not significantly altered (Fig. 9) .

Claims

WE CLAIM:

1. A transgenic mouse wherein expression of the gene encoding amylin is suppressed.

2. The transgenic mouse of claim 1 wherein expression of the gene encoding amylin is decreased as compared to a wild type mouse.

3. The transgenic mouse of claim 2 wherein amylin expression is suppressed by disruption of the gene encoding amylin through insertion into the genome of the mouse a nucleic acid sequence comprising at least a potion of one exon of the amylin encoding sequence linked to a marker sequence.

4. The transgenic mouse of claim 3 wherein said marker sequence is a neomycin resistance gene.

5. The transgenic mouse of claim 3 wherein said marker sequence is a beta galactosidase gene.

6. A transgenic non-human mammalian animal whose germ cells and somatic cells contain an inactivated amylin sequence as a result of chromosomal incorporation into the animal genome, or into the genome of an ancestor of said animal.

7. A transgenic non-human mammalian animal according to claim 6 which is a rodent .

8. A transgenic non-human mammalian animal according to claim 7 wherein said rodent is a mouse.

9. A method of evaluating or identifying an amylin agonist compound, comprising the steps of:

(a) administering said compound to an animal according to any of claims 1-8; and (b) analyzing the animal for amylin activi¬ ty.