Insulin Action and Its Disturbances in Disease - part 1 ppsx

It was shown to bind insulin with high sub-nanomolaraffinity, marked pH dependence decreased affinity even at mildly acid pH andunexpectedly complex kinetics manifested as negative co-oper

Insulin Resistance

Insulin Resistance

Insulin Action and Its Disturbances in Disease

Editors

Sudhesh Kumar

Unit for Diabetes and Metabolism, Warwick Medical School, University of

Warwick, Coventry CV4 7AL, UK

Stephen O’Rahilly

Department of Clinical Biochemistry, University of Cambridge, Addenbrookes

Hospital, Hill Road, Cambridge CB2 2QQ, UK

Copyright  2005 John Wiley & Sons Ltd, The Atrium, Southern Gate, Chichester,

West Sussex PO19 8SQ, England

Email (for orders and customer service enquiries): cs-books@wiley.co.uk

Visit our Home Page on www.wileyeurope.com or www.wiley.com

All Rights Reserved No part of this publication may be reproduced, stored in a retrieval system or transmitted in any form or by any means, electronic, mechanical, photocopying, recording, scanning or otherwise, except under the terms of the Copyright, Designs and Patents Act 1988 or under the terms

of a licence issued by the Copyright Licensing Agency Ltd, 90 Tottenham Court Road, London W1T 4LP, UK, without the permission in writing of the Publisher Requests to the Publisher should be addressed to the Permissions Department, John Wiley & Sons Ltd, The Atrium, Southern Gate,

1243 770620.

This publication is designed to provide accurate and authoritative information in regard to the subject matter covered It is sold on the understanding that the Publisher is not engaged in rendering professional services If professional advice or other expert assistance is required, the services of a competent professional should be sought.

Other Wiley Editorial Offices

John Wiley & Sons Inc., 111 River Street, Hoboken, NJ 07030, USA

Jossey-Bass, 989 Market Street, San Francisco, CA 94103-1741, USA

Wiley-VCH Verlag GmbH, Boschstr 12, D-69469 Weinheim, Germany

John Wiley & Sons Australia Ltd, 33 Park Road, Milton, Queensland 4064, Australia

John Wiley & Sons (Asia) Pte Ltd, 2 Clementi Loop #02-01, Jin Xing Distripark, Singapore 129809 John Wiley & Sons Canada Ltd, 22 Worcester Road, Etobicoke, Ontario, Canada M9W 1L1

Wiley also publishes its books in a variety of electronic formats Some content that appears

in print may not be available in electronic books.

Library of Congress Cataloging-in-Publication Data

Insulin resistance : insulin action and its disturbances in disease / editors, Sudhesh

Kumar, Stephen O’Rahilly.

British Library Cataloguing in Publication Data

A catalogue record for this book is available from the British Library

ISBN 0-470-85008-6

Typeset in 10.5/13pt Times by Laserwords Private Limited, Chennai, India

Printed and bound in Germany

This book is printed on acid-free paper responsibly manufactured from sustainable forestry

in which at least two trees are planted for each one used for paper production.

Ken Siddle

Daniel Konrad, Assaf Rudich and Amira Klip

2.2 Insulin as a master regulator of whole body glucose disposal 63 2.3 Insulin-mediated regulation of glucose metabolic pathways 67 2.4 Glucose uptake into skeletal muscle – the rate-limiting step in glucose

Keith N Frayn and Fredrik Karpe

3.1 Introduction: does insulin affect lipid metabolism? 87 3.2 Molecular mechanisms by which insulin regulates lipid metabolism 88

3.4 Insulin, lipoprotein lipase and cellular fatty acid uptake 94 3.5 Co-ordinated regulation of fatty acid synthesis and ketogenesis 96

vi CONTENTS

Laura J S Greenlund and K Sreekumaran Nair

Gema Medina-Gomez, Christopher Lelliott and

Antonio J Vidal-Puig

5.2 Genetic modification as a tool to dissect the mechanisms leading to insulin

5.3 Candidate genes involved in the mechanisms of insulin resistance 134

5.6 Defining the function of the insulin cascade molecules through global

with Specific Reference to Metabolic Syndrome and Type 2

Stanley M Hileman and Christian Bjørbæk

7.2 Counter-regulation of hypoglycaemia – role of the CNS 180

Philip G McTernan, Aresh Anwar and Sudhesh Kumar

8.5 Obesity and its association with insulin resistance 210

8.7 The pathogenic significance of abdominal adipose tissue 211 8.8 Potential mechanisms linking central obesity to the metabolic syndrome 212 8.9 Randle hypothesis/glucose– fatty acid hypothesis 212

8.15 Visceral obesity and steroid hormone metabolism 217

Robert K Semple and Stephen O’Rahilly

Daniel K Clarke and Vidya Mohamed-Ali

viii CONTENTS

Jeremy Krebs and Susan Jebb

Nicholas J Wareham, Søren Brage, Paul W Franks and

Rebecca A Abbott

12.2 Evidence from observational studies of the association between physical

12.3 Summary of findings from observational studies in adults 318 12.4 Summary of findings from observational studies in children and adolescents 340 12.5 Mechanisms underlying the association between physical activity and insulin

12.6 Trials of the effect of physical activity on insulin sensitivity in adults 353 12.7 Trials of the effect of physical activity on insulin sensitivity in children and

12.8 Evidence of heterogeneity of the effect of physical inactivity on insulin

George Argyropoulos, Steven Smith and Claude Bouchard

Benoˆıt Lamarche and Jean-Fran¸cois Mauger

14.3 Obesity versus the insulin resistance syndrome 453

14.7 LDL cholesterol levels versus LDL particle number 459

CONTENTS ix

14.8 Insulin resistance, dyslipidaemia and the risk of cardiovascular disease 460

Stephen J Cleland and John M C Connell

15.2 Hyperinsulinaemia, insulin resistance and hypertension 467 15.3 Possible mechanisms linking insulin with blood pressure 468

15.5 Vascular endothelial dysfunction and mechanisms of atherothrombotic disease 469

15.7 What causes abnormal insulin signalling in metabolic and vascular tissues? 474

Neus Potau

16.2 Definition of polycystic ovary syndrome (PCOS) and diagnostic criteria 486

16.8 Treatment approach with insulin sensitizers (metformin) 498 16.9 Treatment approach with insulin sensitizers (thiazolidinediones) 501

David Savage and Stephen O’Rahilly

17.2 General biochemical and clinical features of severe insulin resistance 512 17.3 Classification of specific syndromes of insulin resistance 514

17.5 Lipodystrophic syndromes and a lipocentric approach to diabetes 518 17.6 Complex genetic syndromes associated with severe insulin resistance 525 17.7 Therapeutic options in the syndromes of severe insulin resistance 526

Harpal S Randeva, Margaret Clarke and Sudhesh Kumar

x CONTENTS

18.8 Pharmacological treatment of insulin resistance 544 18.9 Insulin sensitizers and cardiovascular risk factors 551

Bei B Zhang and David E Moller

Hormone resistance syndromes are typically thought of as rare, usually genetic,disorders with a severe but relatively stereotyped clinical and biochemical pro-file While there are syndromes of severe insulin resistance that conform tothis description, defective insulin action is of much more pervasive biomedicalimportance Even moderate degrees of insulin resistance are closely linked to

a range of common diseases, including Type 2 diabetes, polycystic ovary drome, obesity and hypertension Not surprisingly, in recent years, there has been

syn-a tremendous incresyn-ase in interest within the medicsyn-al syn-and scientific community

in understanding the causes, consequences and treatment of insulin resistance.There are several reasons for this Firstly, we are now witnessing a revolution

in unravelling the molecular mechanism of insulin action and in understandingthe molecular basis for the various syndromes associated with insulin resistance.Secondly, we are now seeing a global epidemic of Type 2 diabetes that maypose a major threat to international public health Thirdly, the pharmaceuticaland biotechnology industries are investing heavily in the development of newdrugs that can improve insulin action Therefore, we believe that the publication

of this book is timely

There is considerable literature available on the subject of insulin resistance Arecent search on Medline revealed more than 20,000 articles on this subject Thisinformation is readily accessible and one might argue that a book such as thisone might become outdated as soon as it is published! One guiding principle forthis book was, therefore, to bring to the reader not only a synthesis of importantinformation, but also the wisdom of leading researchers and clinicians who arerecognised as leaders in their own fields

Each chapter stands independently and is written by one or more experts onthe subject The book is divided into five sections with a total of 19 chapters.Section 1 reviews our current understanding of the normal biology of insulinaction and separate chapters cover insulin action in relation to glucose, lipid andprotein metabolism Section 2 explores the pathophysiological mechanisms ofinsulin resistance, with discussion of the effects of glucose disposal in humansand in animal models It also reviews the central regulation of energy metabolismand its perturbation, as well as the relationship between fat distribution andinsulin action and the role of the nuclear hormone receptor PPARγ in glucosemetabolism Finally, there is a chapter discussing the role of adipose tissue-secreted products in causing insulin resistance Section 3 examines the role of

Although the book aims to provide comprehensive coverage of the subject,there are some obvious omissions, for example, the relationship between insulinresistance and Type 2 diabetes Whilst this relationship is alluded to in manyplaces, we have not devoted a full chapter to it as there are several excellentrecent reviews on the subject.

The book is intended mainly for a specialist readership, although it may prove

to be a useful resource for a wide variety of scientists, clinicians and postgraduatestudents with an interest in any of the related conditions We hope that regardless

of your background as a physician, medical researcher or scientist, you willfind this book appropriate for your needs Finally, all contributing authors haveproduced outstanding chapters that reflect their expertise and wisdom and sparedtheir valuable time despite tremendous pressures from competing obligations

We wish to thank them all for their support, hard work and friendship

Sudhesh Kumar Stephen O’Rahilly

August 2004

List of Contributors

Rebecca A Abbott, MRC Epidemiology Unit, Strangeways Research

Laboratory, Worts Causeway, Cambridge CB1 8RN, UK

Frank Alford, St Vincent’s Hospital, Melbourne, Endocrine Unit, 41 Fitzroy

Parade, Fitzroy, Victoria 3065, Australia

Aresh Anwar, University Hospitals of Coventry and Warwickshire, Walsgrave

Hospital, Clifford Bridge Road, Coventry CV2 2DX, UK

George Argyropoulos, Pennington Biomedical Research Center, 6400 Perkins

Road, Baton Rouge, LA 70808, USA

Henning Beck-Nielsen, Odense University Hospital, Department of

Endocrinology, Kloevervaenget 64, 5000 Odense C, Denmark

Christian Bjørbæk, Division of Endocrinology, Beth Israel Deaconess

Medical Center Research North, 330 Brookline Avenue, Boston, MA

02215, USA

Claude Bouchard, Pennington Biomedical Research Center, 6400 Perkins

Road, Baton Rouge, LA 70808, USA

Søren Brage, MRC Epidemiology Unit, Strangeways Research Laboratory,

Worts Causeway, Cambridge CB1 8RN, UK

Daniel K Clarke, Adipokines and Metabolism Research Group, Department

of Medicine, University College London, 48 Riding House Street, LondonW1W 7EY, UK

Margaret Clarke, Heartlands and Solihull NHS Trust, Birmingham B19

9RA, UK

Stephen J Cleland, Department of Medicine and Therapeutics, University of

Glasgow, Glasgow G11 6NT, UK

John M C Connell, Division of Cardiovascular and Medical Sciences,

University of Glasgow, Glasgow G11 6NT, UK

Paul W Franks, MRC Epidemiology Unit, Strangeways Research

Laboratory, Worts Causeway, Cambridge CB1 8RN, UK

xiv LIST OF CONTRIBUTORS

Keith N Frayn, Oxford Centre for Diabetes, Endocrinology and Metabolism,

Churchill Hospital, Oxford OX3 7LJ, UK

Laura J S Greenlund, Department of Endocrinology, Mayo Clinic, 200 First

Street SW, Rochester, MN 55905, USA

Stanley M Hileman, Department of Physiology and Pharmacology, West

Virginia University, Morgantown, WV 26506, USA

Ole Hother-Nielsen, Odense University Hospital, Department of

Endocrinology, Sdr Boulevard 29, 5000 Odense C, Denmark

Susan Jebb, MRC Human Nutrition Research, Elsie Widdowson Laboratory,

Fulbourn Road, Cambridge CB1 9NL, UK

Fredrik Karpe, Oxford Centre for Diabetes, Endocrinology and Metabolism,

Churchill Hospital, Oxford OX3 7LJ, UK

Amira Klip, Programme in Cell Biology, The Hospital for Sick Children, 555

University Avenue, Toronto, ON, M5G 1X8, Canada

Daniel Konrad, Programme in Cell Biology, The Hospital for Sick Children,

555 University Avenue, Toronto, ON, M5G 1X8, Canada

Jeremy Krebs, Wellington Clinical School of Medicine, University of Otago,

P.O Box 7343, Wellington South, New Zealand

Sudhesh Kumar, Unit for Diabetes and Metabolism, Warwick Medical

School, University of Warwick, Coventry CV4 7AL, UK

Benoˆıt Lamarche, Institute on Nutraceuticals and Functional Foods, 2440

Boulevard Hochelaga, Laval University, Quebec, G1K 7P4, Canada

Christopher Lelliott, Department of Clinical Biochemistry and Metabolic

Medicine, University of Cambridge, Addenbrooke’s Hospital, Hills Road,Cambridge CB2 2QR, UK

Jean-Fran¸cois Mauger, Institute on Nutraceuticals and Functional Foods,

2440 Boulevard Hochelaga, Laval University, Quebec, G1K 7P4, Canada

Philip G McTernan, Unit for Diabetes and Metabolism, Warwick Medical

School, University of Warwick, Coventry CV4 7AL, UK

Gema Medina-Gomez, Department of Clinical Biochemistry and Metabolic

Medicine, University of Cambridge, Addenbrooke’s Hospital, Hills Road,Cambridge CB2 2QR, UK

Vidya Mohamed-Ali, Adipokines and Metabolism Research Group,

Department of Medicine, University College London, 48 Riding House Street,London W1W 7EY, UK

LIST OF CONTRIBUTORS xv David E Moller, Departments of Molecular Endocrinology and Metabolic

Disorders, Merck Research Laboratories, Rahway, NJ 07065, USA

K Sreekumaran Nair, Department of Endocrinology, Mayo Clinic, 200 First

Street SW, Rochester, MN 55905, USA

Stephen O’Rahilly, Department of Clinical Biochemistry and Medicine,

University of Cambridge, Addenbrooke’s Hospital, Hills Road, CambridgeCB2 2QQ, UK

Neus Potau, Hormonal Laboratory, Hospital Matemo-Infantil Vail d’Hebron,

Passeig Vail d’Hebron, 119–129, 08035 Barcelona, Spain

Harpal S Randeva, Molecular Medicine Research Group, Biomedical

Research Institute, Biological Sciences, University of Warwick, CV4 7AL, UK

Assaf Rudich, Programme in Cell Biology, The Hospital for Sick Children,

555 University Avenue, Toronto, ON, M5G 1X8, Canada

David Savage, Department of Clinical Biochemistry and Medicine, University

of Cambridge, Addenbrooke’s Hospital, Hills Road, Cambridge CB2 2QQ, UK

Robert K Semple, Department of Clinical Biochemistry, University of

Cambridge, Addenbrooke’s Hospital, Hills Road, Cambridge CB2 2QR, UK

Ken Siddle, Department of Clinical Biochemistry, University of Cambridge,

Addenbrooke’s Hospital (Box 232), Hills Road, Cambridge CB2 2QR, UK

Stephen Smith, Pennington Biomedical Research Center, 6400 Perkins Road,

Baton Rouge, LA 70808, USA

Antonio J Vidal-Puig, Department of Clinical Biochemistry and Metabolic

Medicine, University of Cambridge, Addenbrooke’s Hospital, Hills Road,Cambridge CB2 2QR, UK

Nicholas J Wareham, MRC Epidemiology Unit, Strangeways Research

Laboratory, Worts Causeway, Cambridge CB1 8RN, UK

Bei B Zhang, R80W180, Merck Research Laboratories, P.O Box 2000, 126

E Lincoln Avenue, Rahway, NJ 07065, USA

The Insulin Receptor

and Downstream Signalling

Ken Siddle

Insulin regulates diverse physiological processes in mammals, including brane transport, intermediary metabolism and cell growth and differentiation.These actions involve rapid effects on subcellular membrane traffic, enzymeactivity and protein synthesis as well as longer term actions on gene tran-scription The most conspicuous metabolic effects of insulin are associatedwith skeletal muscle, adipose tissue and liver1 but its physiologically impor-tant actions are by no means confined to such tissues, as evidenced by thephenotypes of mice with tissue-specific knockout of insulin receptor in brain,pancreatic β-cells or endothelia.2 Insulin signalling pathways have also beenimplicated in accelerating the ageing process.3, 4

mem-Understanding of the signalling pathways by which the insulin receptor isable to influence so many and such diverse cellular targets is still far fromcomplete, although the last 20 years have seen major advances A surprisingfeature is that to date the only signalling component known to be unique toinsulin action is the insulin receptor (IR) itself, which is widely expressed inmammalian cells, although levels vary greatly between cell types The IR bindsinsulin with high affinity and specificity, and transmits a signal to the cytosolvia its intrinsic tyrosine-specific protein kinase activity This phosphorylates anumber of intracellular substrates, most especially the so-called insulin receptorsubstrates (IRSs), which recruit and activate an array of signalling proteins con-taining Src homology-2 (SH2) domains Two signals have been shown to playmajor roles in insulin action, namely those transmitted by the enzyme phospho-inositide 3-kinase (PI 3-kinase), which generates PtdIns(3,4,5)tris-phosphate at

Insulin Resistance. Edited by Sudhesh Kumar and Stephen O’Rahilly

 2005 John Wiley & Sons, Ltd ISBN: 0-470-85008-6

2 THE INSULIN RECEPTOR AND DOWNSTREAM SIGNALLING

the cytosolic face of membranes, and the guanine nucleotide exchange factorGrb2/Sos, which activates the small G-protein Ras These act as switch mech-anisms to change the ‘currency’ of signalling from tyrosine phosphorylation toserine/threonine phosphorylation of target proteins However, these signals, andthe downstream signalling cascades involving protein kinase B and mitogen-activated protein kinases (MAPKs), have been implicated in the actions of awide variety of hormones and growth factors as well as specific actions ofinsulin This chapter will focus particularly on the IR and its substrates, andconsider more briefly what is known about downstream signalling pathways,which have been reviewed in detail elsewhere.5 – 9

The insulin receptor family

The IR is a large, heterotetrameric, transmembrane glycoprotein containing twotypes of subunit, designated α (Mr 140 kDa) and β (95 kDa), linked by disul-phide bonds in a β–α–α–β configuration The principal members of the IRfamily of receptor tyrosine kinases are represented in Figure 1.1, together withtheir high affinity ligands It is possible that IRR may also form hybrids with IR,although because of the very restricted distribution of IRR these are unlikely to

be of major significance It is likely that the two isoforms of IR will also formheterodimers, although this has recently been questioned.22 It is assembled from

a single polypeptide pro-receptor, by dimerization, proteolytic cleavage and cosylation within the endoplasmic reticulum and Golgi apparatus, before traffick-ing of mature receptor to the plasma membrane The IR was initially defined byradioligand binding studies, which provided information on affinity, specificityand tissue distribution It was shown to bind insulin with high (sub-nanomolar)affinity, marked pH dependence (decreased affinity even at mildly acid pH) andunexpectedly complex kinetics (manifested as negative co-operativity).10 Thefirst real insight into signalling mechanisms came with the demonstration thatthe receptor possessed intrinsic, tyrosine-specific protein kinase activity that wasstimulated by insulin binding.11 Soon afterwards, cloning of the pro-receptorcDNA12, 13 and the receptor gene14 opened the door to analysis of receptorstructure and function, which is now understood in considerable detail.15 – 18The IR gene consists of 22 coding exons spanning 120 kilobases on chro-mosome 19p13.2 Exon 11, of just 36 nts, is subject to alternative splicing,resulting in the generation of two isoforms designated IR-A (Ex 11−) and IR-B(Ex 11+), which differ in sequence by 12 amino acids at the carboxyl-terminus

gly-of theα-subunit (the numbering used here includes the exon 11 sequence) Therelative proportions of the two isoforms differ between tissues, IR-A predom-inating in brain and IR-B in liver, while both are found in similar amounts inskeletal muscle and placenta.19, 20The isoforms differ modestly but significantly

INSULIN RECEPTOR STRUCTURE AND FUNCTION 3

4 THE INSULIN RECEPTOR AND DOWNSTREAM SIGNALLING

with respect to their binding affinity for insulin and IGFs, and this is perhaps notsurprising in view of the proximity of the variable sequence to a known majorbinding epitope at theα-subunit carboxyl-terminus More controversially, it hasbeen suggested that the short peptide sequence encoded by exon 11 also acts as

a sorting signal, causing the isoforms to localize to different plasma membranemicrodomains from which they activate distinct signalling cascades.21, 22The IRgene is transcribed as mRNAs of 7–11 kilobases, which include substantial 5
-and 3
-untranslated regions either side of the coding sequence (NCBI databasereferences for complete IR (InsR) cDNA and protein sequences are human NM

000208, mouse NM 010568 and rat NM 017071) The deduced sequence of thehuman IR precursor contains 1382 (or 1370) amino acids, including a signalsequence of 27 amino acids which is absent from mature receptor A tetrabasicRKRR motif marks the site of proteolytic cleavage to generate the α- and β-subunits, of 731 (719) and 620 amino acids respectively In the mature receptortheα-subunit is wholly extracellular and contains the ligand binding site, whilethe β-subunit contains a single predicted membrane-spanning segment and anintracellular tyrosine kinase domain

The extracellular portion of IR is heavily glycosylated, and some tion is essential for normal receptor function.23 Disulphides betweenα-subunitsare contributed by Cys524 and Cys682/3/5,24and can readily be reduced in vitro

glycosyla-to generate half-recepglycosyla-tors that bind insulin with decreased affinity In contrastthe α–β disulphide between Cys647 and Cys872 can be reduced only underdenaturing conditions Experimental perturbation of glycosylation,24 proteolyticcleavage25 and disulphide bonding26 can profoundly affect receptor function.There is no compelling evidence that these processes are modulated under

physiological conditions in vivo, although it remains possible that there are

cir-cumstances where this does occur

Just as insulin is structurally related to the insulin-like growth factors, sothe IR is similar in structure and function to the type 1 IGF receptor (IGFR),with which it shares approx 60 per cent amino acid sequence identity.27 (Thetype 2 IGF receptor is an unrelated protein,28 which is not thought to have anysignalling function but may have a role in clearance of IGF from the circula-tion.) Like the IR, the IGFR is very widely expressed, albeit at different levels.There is significant expression of IGFR in skeletal muscle, but very low levels

in hepatocytes and adipocytes The very different biological roles of IR andIGFR are emphasized by the distinct phenotypes of mouse knockout models.Mice lacking IR exhibit only slight (10 per cent) growth retardation at birth, butdie within days as a result of uncontrolled hyperglycaemia and ketoacidosis,2

although lack of IR causes more severe growth retardation in humans In trast, mice lacking IGFR are severely growth deficient (approximately 45 percent of normal size) and developmentally retarded and die at birth of respira-tory failure.29 However, the functions of the two receptors are not completely

con-INSULIN RECEPTOR STRUCTURE AND FUNCTION 5

distinct as shown by the efficacy of IGF-I in reducing hyperglycaemia in humansubjects lacking functional IR.30

A third member of the IR/IGFR family is the insulin-receptor-related receptor(IRR).31Although this has a similar degree of homology to IR and IGFR as thesereceptors do to each other, the IRR does not bind either insulin or IGFs,32and noligand has yet been identified for this receptor Expression of IRR is much morerestricted than that of IR and IGFR, but it is found in kidney, neural tissue, stom-ach and pancreatic beta cells Mice lacking IRR appear phenotypically normal,33although there is evidence that, along with other members of the IR family, IRRcontributes in a non-redundant fashion to testicular development in mice.34Insulin binds to IGFR with low affinity, which would not be sufficient to

permit significant activation by insulin in vivo under normal physiological

condi-tions, but could become important under pathological conditions associated withhyperinsulinaemia Indeed, it has been suggested that such ‘specificity spillover’might contribute to features of insulin resistance syndromes such as acanthosisnigricans and polycystic ovaries35, 36and it may well be responsible for effects

of insulin on growth of cultured cells The converse phenomenon, stimulation

of the IR by IGFs, may be of greater physiological significance Early studies ofligand specificity, which gave rise to the notion that IGFs bound only with lowaffinity to IR, commonly used rat liver as a source of receptors and such studiesreflected the properties of the IR-B isoform In fact, although the isoforms differonly slightly in affinity for insulin itself, the A isoform has substantially higheraffinity for IGFs, particularly IGF-II, than the B-isoform.37, 38 Indeed the affin-ity of IR-A for IGF-II is comparable to that of the type 1 IGFR, and it appearsthat IR-A makes a significant contribution to mediating biological activity of

IGF-II, both in vivo and in vitro.39, 40

When IR and IGFR are expressed in the same cells, they are can form hybridstructures containing an insulin half-receptor, disulphide linked to an IGF half-receptor (Figure 1.1).41 – 43 Surprisingly heterodimerization of proreceptors toform hybrids seems to occur with similar efficiency to homodimerization to formclassical receptors, so the proportion of receptors existing as hybrids is largely

a reflection of the relative expression levels of the individual receptors.43 – 45

Hybrid receptors thus occur commonly in vivo, and in tissues such as heart

and skeletal muscle, where IR is expressed at higher levels than IGFR, hybridsaccount for the majority of high affinity ‘IGF receptors’.44 Conversely, whenIGFR is in excess, as in fibroblasts, the majority of IR is drawn into hybrids

It remains possible that mechanisms exist that promote or inhibit assembly ofhybrid receptors but these have not been demonstrated Hybrid receptors bindIGF with high affinity, comparable to classical type 1 IGFR, and would there-

fore be expected to contribute significantly to mediating IGF actions in vivo.

However, hybrids bind insulin with relatively low affinity, especially those porating the IR-B isoform,46, 47 and are unlikely to contribute significantly to

incor-6 THE INSULIN RECEPTOR AND DOWNSTREAM SIGNALLING

insulin signalling at physiological insulin concentrations In fact if IR are porated into hybrids this would be expected to decrease cellular sensitivity toinsulin, and it has been suggested that an increase in the proportion of hybridreceptors in skeletal muscle of obese and diabetic subjects may contribute toinsulin resistance.48, 49 However, in classical insulin target tissues such as liver,

incor-fat and muscle the proportion of IR in hybrids is always likely to be small andthe potential for changes in IGFR expression to influence insulin sensitivity must

be correspondingly slight It is unclear whether hybrid receptors have uniquesignalling properties, which might influence the nature of cellular responses Itwould be expected that binding of either insulin or IGF would lead to activation

of tyrosine kinase activity within both β-subunits of hybrids.50 As discussedbelow, the signalling competencies of IR and IGFR are very similar but proba-bly not identical In this context, hybrids might in principle have the signallingproperties of both IR and IGFR, or even additional novel properties reflectingsynergy between the individual half-receptors

The IR extracellular domain: ligand binding

Apart from the intrinsic interest of unravelling the molecular basis of and binding and the mechanism of receptor activation, understanding of lig-and–receptor interactions could facilitate the design of insulin mimetics withtherapeutic potential However, the large size of the IR has presented a con-siderable analytical challenge Molecular modelling based on sequence analysispredicts that the extracellular portion of each half-receptor contains six dis-tinct structural domains, while three intracellular domains are recognized.51

lig-The N-terminal, membrane-distal half of the extracellular receptor containstwoβ-helical L domains flanking a cysteine-rich (CR) region (Figure 1.2) Thestructural domains of the IR are shown in Figure 1.2: L1 and L2 are β-helicaldomains; CR is the cysteine-rich domain; Fn0, Fn1 and Fn2 are fibronectin typeIII repeats; the inserted region within Fn1 includes the site of cleavage betweenα- and β-subunits; JM is the juxtamembrane region; TK is the tyrosine kinasedomain; CT is the carboxyl-terminal domain Positions of inter-subunit disul-phide links and ligand binding epitopes are as indicated The correspondingportion of the IGFR, expressed as a recombinant protein, has been crystal-lized and its structure has been determined.52 This reveals the L domains anddisulphide-bonded modules of the CR domain surrounding a putative ligand-binding cavity (although this IGFR fragment does not itself bind IGFs) Theorientation of the L domains within the crystal may not be the same as in nativereceptor, and of course differences in conformation between IR and IGFR mightcontribute to binding specificity However, it is safe to assume that the struc-tures of the L1/CR/L2 domains of the IR are similar to those of the IGFR Theremaining extracellular portion of both IR and IGFR is believed to consist ofthree fibronectin type III domains, each folded as a seven-stranded β-sandwich

INSULIN RECEPTOR STRUCTURE AND FUNCTION 7

8 THE INSULIN RECEPTOR AND DOWNSTREAM SIGNALLING

However, the limitations of theoretical modelling are illustrated by the factthat different groups who predicted structures for the first such domain (usu-ally referred to as the Fn0 domain) assigned differentβ-strands within the FnIIIfold,53, 54 and therefore proposed different positions for the inter-subunit disul-

phide bond formed by Cys524 The central FnIII domain (usually referred to asFn1) contains a large inserted region as a loop betweenβ-strands, for which noparticular secondary or tertiary structure is predicted In the middle of this arethe sequence encoded by the alternatively spliced Exon 11 and the site of cleav-age between α- and β-subunits The Fn1 domain is therefore assembled fromsequences within the C-terminal region of the α-subunit and the N-terminalregion of theβ-subunit, so that the α- and β-subunits are not readily dissociated

as independent proteins but are held together by strong non-covalent forces aswell as disulphide bonds

The residues in insulin that are important for receptor binding have been sively studied by comparing the properties of insulins from different species,

inten-by chemical modification and inten-by mutational analysis and alanine scanning.10, 55

These studies indicate that two surfaces of the insulin molecule are tant for receptor binding.15The ‘classical’ binding surface contains a number ofhydrophobic residues (including B24Phe and B25Phe), while the second bindingsurface is more polar Both surfaces are essentially ‘conformational’ in natureand include residues from disparate regions of primary sequence A screen oflarge, random, phage-displayed peptide libraries has identified novel peptidesthat bind to the IR at or close to the insulin binding site and presumably mimiccritical aspects of the insulin surface.56 Indeed derivatives of these peptidesfunction as insulin mimetics and activate the receptor.57 It remains to be seenwhether it will be possible to model non-peptide mimetics using informationderived from the study of such peptides

impor-The task of identifying residues in the IR/IGFR that contact ligand is moredifficult, given the very large size of the receptors The problem has beenapproached by cross-linking insulin analogues, constructing IR/IGFR chimeras,and by mutational analysis and alanine scanning (as reviewed in 9, 15 and16) Four distinct binding epitopes have been identified, within the L1, CR, L2and Fn1 insert domains Residues in the L1 and L2 domains of IR, especiallyPhe39, are important for insulin specificity, although IGFR specificity for IGF-1

is more dependent on residues in the CR domain These putative binding topes flank the cavity enclosed by the L1, CR and L2 domains in the crystalstructure of the IGFR fragment described above, and this cavity is of appro-priate dimensions to accommodate a molecule of ligand.16 Although neitherthis IGFR fragment nor the corresponding IR fragment bind ligand, addition toeither construct of the fourth binding epitope, a short peptide sequence fromtheα-subunit carboxyl-terminus, confers ligand binding of moderate affinity.58

epi-Remarkably, this peptide confers binding ability on N-terminal fragments notonly when fused directly but even when added as a free peptide.59 Mutational

INSULIN RECEPTOR STRUCTURE AND FUNCTION 9

analysis also shows that this sequence, and particularly two Phe residues within

it, makes a major contribution to binding affinity without influencing ficity for insulin versus IGF-I.60 Cross-linking studies show that this sequence

speci-is in proximity to the PheB25 of bound insulin, suggesting that hydrophobicinteractions between these regions provide a major part of the binding energy.Remarkably, the nearby PheB29 lies close to the L1 domain of the IR,15 indi-cating that the N- and C-terminal domains of the receptorα-subunit are closelyjuxtaposed within the native structure It is not yet clear how individual bindingepitopes are assembled in three dimensions to create the high affinity, negativelyco-operative insulin binding characteristic of native IR Highest affinity binding

is seen only in the context of dimeric constructs, and the preferred model ofligand binding is one in which both α-subunits contribute asymmetrically tothe insulin binding site, and a single molecule of bound insulin contacts bothhalves of the receptor15 (Figure 1.3) Figure 1.3 shows a hypothetical model ofinsulin binding to IR, as viewed from perpendicular to the extracellular mem-brane face The twoα-subunits are aligned antiparallel, with binding epitopes e1and e2 contributed by L1 and L2 domains respectively (other binding epitopesare not shown) The model is such that only a single molecule of insulin bindswith high affinity, and cross-links the α-subunits Such a model is compatiblewith many observations concerning the structural requirements and kinetics ofinsulin binding, not least the simple fact that, in spite of its dimeric structure,the IR binds only a single molecule of insulin with high affinity Electron micro-scopic images of gold-labelled insulin bound to the receptor are also broadlyconsistent with this model.17 A precedent for such a binding mechanism exists

in the complex of growth hormone with its receptor.61 However, crystallization

of the EGFR in complex with ligand reveals a different binding mechanism,

in which L1/CR/L2 receptor domains dimerize back to back and each bind amolecule of ligand.62, 63 Confirmation of the insulin binding mechanism must

await crystallization of insulin–IR complexes

The IR intracellular domain: tyrosine kinase activation

and autophosphorylation

Ligand binding induces conformational changes in the extracellular portion

of the receptor, which in turn must alter the conformation of, or relationshipbetween, the intracellular domains in a manner that promotes autophosphory-lation The activation of the tyrosine kinase domains depends largely on recip-

rocal intramolecular trans-phosphorylation between β-subunits.50, 64, 65 Both

strict intramolecular phosphorylation (within β-subunits)66 and intermolecularphosphorylation (between hetero-tetrameric receptors)67 have also been demon-strated, although the latter appears not to be sufficient to stimulate substratekinase activity.68

10 THE INSULIN RECEPTOR AND DOWNSTREAM SIGNALLING

INSULIN RECEPTOR STRUCTURE AND FUNCTION 11

Within the intracellular portion of the IR, the tyrosine kinase domain proper,

of approximately 250 amino acids, is flanked by a juxtamembrane (JM) domain

of approximately 50 amino acids and carboxyl-terminal (CT) domain of imately 100 amino acids (Figure 1.4) The principal phosphorylation sites areindicated in Figure 1.4, with their functional significance where known, togetherwith the critical lysine 1030 required for ATP binding and catalytic activity Thecrystal structure of the kinase domain has been determined.18 There are sites oftyrosine autophosphorylation in all three domains, as well as multiple potentialsites of serine phosphorylation.69However, while tyrosine phosphorylation siteshave been well defined, the extent and significance of serine phosphorylationremains unclear The IR tyrosine kinase domain has been crystallized in bothbasal, unphosphorylated, and activated, phosphorylated, states, and this has pro-vided important insights into the mechanisms of catalysis and regulation.18 Thebi-lobed structure is broadly typical of other protein kinases The active sitelies in a cleft between the two lobes and includes Lys1030 and other residuesimportant in ATP binding The size and hydrophobicity of this cleft confersspecificity for phosphorylation of tyrosine rather than serine residues In thebasal state the cleft is effectively closed by a regulatory peptide loop, and theactive site is inaccessible to peptide substrates Autophosphorylation of thisloop, on tyrosines 1158, 1162 and 1163, causes it to swing away from the cleft,allowing access of other substrates to the active site The IGF receptor tyrosinekinase has a very similar structure and activation mechanism.70, 71

approx-In addition to its critical role in tyrosine kinase activation, tion also facilitates recruitment of substrates and adaptor proteins Phosphory-lation of a conserved NPEY motif in the JM domain of IR and IGFR creates

autophosphoryla-a binding site for the PTB (phosphotyrosine-binding) domautophosphoryla-ains of insulin tor substrates (IRSs) and Shc proteins.72, 73 IRS-2 additionally has a region(the KRLB domain) that binds directly to the phosphorylated kinase regulatoryloop.74, 75 The substrate APS (adaptor with PH and SH2 domains) also interacts,via its SH2 domain, with phosphotyrosine residues of the activation loop,76 asdoes the non-substrate adaptor Grb10.77 The CT domain of IR contains twosites of autophosphorylation, Y1328 and Y1334, of which only the latter is con-served in IGFR These sites bind a number of SH2 domain-containing adaptors

recep-in vitro78, 79 although the contribution to insulin signalling in vivo may be small.

Mutation of these sites or deletion of a larger segment of the CT domain hasbeen reported to influence signalling specificity, and particularly to affect therelative efficiency of metabolic versus mitogenic signalling by the IR, althoughthis has not been a consistent finding (reviewed in reference 69) The role ofthese sites in IR function thus remains unclear

Autophosphorylation also acts as a trigger for internalization of the vated receptor/insulin complex,80 which is important both in terminating theinsulin signal and in insulin degradation (this being the major mechanism bywhich insulin is cleared from the circulation, particularly by the liver) The

acti-12 THE INSULIN RECEPTOR AND DOWNSTREAM SIGNALLING

Adaptor binding (Grb10) Adaptor binding ??

INSULIN RECEPTOR STRUCTURE AND FUNCTION 13

endocytic machinery recognizes β-turn tyrosine motifs and/or dileucine motifs

in the receptor juxtamembrane domain.81 – 83 Phosphorylation of the brane tyrosines is not required for internalization, but ligand binding and/orautophosphorylation at other sites causes conformational change that exposesthese motifs, triggering movement of receptor into clathrin-coated pits Coatedvesicles deliver receptor to early endosomes, where acidification causes rapiddissociation of bound insulin, which is then degraded by specific endosomaland/or cytosolic proteases.84, 85 Endosomal IR may contribute transiently but

juxtamem-significantly to signalling, both in terms of prolongation of signal and access

to intracellular substrates.85 Indeed, phosphorylation of Shc (but not IRSs) isdependent to some extent on receptor internalization.86 – 88 Once the stimulus

of bound ligand is removed the action of phosphotyrosine phosphatases results

in rapid inactivation and receptor is largely recycled to the plasma membrane

Several different phosphatases act on IR in vitro, but PTP1B is of particular important in vivo.89, 90Indeed, there is considerable interest in PTP1B as a drugtarget for treatment of diabetes and obesity.91 – 93

The IR is phosphorylated on serine/threonine as well as tyrosine residues,both in response to stimulation by insulin itself and as a result of cross-talkfrom other signalling pathways, and it has been suggested that such phos-phorylation is inhibitory to IR signalling Multiple sites of phosphorylationhave been identified (see for example references 94–97), and phosphorylation

by protein kinase C isoforms has been associated with inhibition of receptorfunction.96, 98–100 It has been suggested that serine/threonine phosphorylation

of IR might mediate glucose-induced inhibition of insulin signalling101, 102 andcontribute to insulin resistance associated with obesity103 and polycystic ovarysyndrome.104 However, it has proved difficult to establish a link between inhi-bition of function and phosphorylation of specific sites,69, 105 and the kinases

responsible for IR phosphorylation in vivo have not been well defined

Phospho-rylated IGFR binds 14–3–3 proteins,106, 107 but this is probably dependent onthe IGFR-specific cluster serines 1280/81/82/83, and no evidence has been pre-sented for a comparable interaction of 14–3–3 with IR Overall, serine/threoninephosphorylation of IR is a somewhat neglected field, in which final conclusionsremain to be drawn

Regulation of insulin receptor expression

The level of IR expression is an important factor determining sensitivity ofcells to insulin, and is probably dependent on controls operating at the level

of both transcription and translation of mRNA The promoter region of the

IR gene has been characterized and binding sites have been identified for anumber of transcription factors Such studies have provided some insights intomechanisms that may be responsible for the almost ubiquitous expression of

14 THE INSULIN RECEPTOR AND DOWNSTREAM SIGNALLING

regulatory effects of hormones,116, 117 and increased expression in some cancercells.118, 119 However, understanding of transcriptional regulatory mechanismsthat underlie the wide divergence in levels of receptor expression in differentcell types, or the modulation of receptor expression under different physiologicalconditions, is still far from complete

The mRNAs encoding both the IR and IGFR have unusually long and tially structured 5’UTRs These may inhibit the normal cap-dependent scanningmechanism of mRNA translation, and/or contain internal ribosome entry sites(IRESs), thus creating the potential for regulation of protein expression at atranslational level through the involvement of additional factors not requiredfor bulk protein synthesis The 5’UTR of the IGFR mRNA has been shown tocontain a functional IRES,120 and it is very likely that this is also the case forthe IR mRNA It remains to be explored whether expression of either receptor

poten-is modulated at a translational level under physiological conditions

The level of receptor expression must also depend on its rate of degradation

As outlined above, activated insulin/receptor complexes are rapidly ized through clathrin-coated pits and delivered to an endosomal compartment.The receptors largely recycle back to the plasma membrane, while insulin isdegraded.80 It has long been known that when cells are exposed to high con-centrations of insulin for long periods receptor degradation is accelerated andexpression is down-regulated,121but IR degradation remains poorly understood

internal-It is possible that ubiquitination is involved as has been shown to be the casewith other receptors,122, 123 and this might be influenced by adapter proteins

that bind to activated IR Thus it has been proposed that both APS (throughassociation with c-Cbl124) or Grb10 (through association with NEDD4125) maymediate receptor ubiquitination, as well as modulating signalling in other ways.There is no evidence that binding affinity of IR is susceptible to direct reg-

ulation in a way that might affect insulin sensitivity in vivo Certainly, affinity

is not thought to be influenced by intracellular phosphorylation events Affinityregulators have from time to time been proposed126but have not been well char-acterized or shown to be physiologically important Likewise, the significance

of MHC class I molecules127 or membrane glycoprotein PC-1128 as tors of IR function remains uncertain In principle, binding affinity might beindirectly influenced by alternative mRNA splicing events affecting the propor-tions of IR-A and B isoforms Data concerning expression of IR isoforms inobesity or diabetes have been inconsistent but in general have not shown signif-icant changes.20, 129–133 On the other hand, aberrant regulation of IR alternative

modula-splicing has been associated with insulin resistance in myotonic dystrophy.134

Changes in IGFR expression would also be expected indirectly to affect insulinsensitivity, by altering the proportion of IR in hybrids, which bind insulin withlower affinity than IR.47 There is evidence that the level of hybrids is increased

in skeletal muscle of obese and diabetic subjects,48, 49 but the hybrid fraction