Diabetes and Vascular Disease Research http://dvr.sagepub.

com/

The application of proteomics to diabetes

Eleanor M Scott, Angela M Carter and John BC Findlay Diabetes and Vascular Disease Research 2005 2: 54 DOI: 10.3132/dvdr.2005.009 The online version of this article can be found at: http://dvr.sagepub.com/content/2/2/54

Published by:
http://www.sagepublications.com

Additional services and information for Diabetes and Vascular Disease Research can be found at: Email Alerts: http://dvr.sagepub.com/cgi/alerts Subscriptions: http://dvr.sagepub.com/subscriptions Reprints: http://www.sagepub.com/journalsReprints.nav Permissions: http://www.sagepub.com/journalsPermissions.nav Citations: http://dvr.sagepub.com/content/2/2/54.refs.html

>> Version of Record - May 1, 2005 What is This?

Downloaded from dvr.sagepub.com at MAHATMA GANDHI INST MED SCI on April 4, 2013

REVIEW

The application of proteomics to diabetes

ELEANOR M SCOTT, ANGELA M CARTER, JOHN BC FINDLAY
Abstract roteomics is the investigation of all the proteins and their various modifications making up a system, be that a cell, tissue or organism. The techniques involved in proteomics allow the global screening of complex samples of proteins and provide qualitative and quantitative evidence of altered protein expression. This lends itself to the investigation of the molecular mechanisms underpinning disease processes and the effects of treatment. This review describes the main techniques of proteomics and how they have begun to be applied to diabetes research. Diabetes Vasc Dis Res 2005;2:5460 In the wider context of insulin resistance, dyslipidaemia is an important component, involving adipose tissue and the liver in its pathogenesis and expression. Ultimately, the main clinical manifestation of diabetes is vascular disease:7 macrovascular ischaemic heart disease and stroke are the main causes of death, dramatically reducing life expectancy, and microvascular complications (retinopathy, nephropathy and neuropathy) are a major cause of morbidity.3 The field of diabetes research has expanded as the laboratory and bioinformatics tools for unravelling complex phenotypes have evolved. Proteomics is the latest research tool to be employed in this context, and it perhaps holds more promise than the genetic analyses that have been prevalent for the past decade or so. The aim of this review article is to provide the reader with an overview of proteomic techniques and their potential application to the study of diabetes. What is proteomics? In a manner analogous to the terminology of genetic studies, the term proteome is used to describe the entire protein complement of a system and proteomics is the study of this complement.8 Proteins are the principal mediators of every cellular process, and the proteome of an organism is far more complex than its genome. Consequently, it is more informative. The genome remains stable throughout life and the genome isolated from one cell type is (for the most part) identical to that isolated from another. Thus, the genome cannot provide information regarding the impact of environmental and other stimuli on a particular organism or cell. Although each of our cells contains all the genes required to make a complete human being, not all of these genes are expressed in every cell or at all times. Therefore, whilst any given organism has one genome, by the nature of selective expression of the genes in different cells and under different stimuli, it can be considered to have many proteomes. To complicate matters further, a protein coded for by a single gene can occur in several different forms as a consequence of alternative splicing of mRNA transcripts and posttranslational modifications (PTMs), see figure 1. At least 200 different PTMs have been identified, including glycosylation, phosphorylation and non-enzymatic glycation (see Delta Mass, http://www.abrf.org/index.cfm/dm.home, a database of protein post-translational modifications). A quick search using the word proteomics on Medline is revealing, with obvious exponential growth in the publication of articles utilising the technology of proteomics, particularly in the last year. By characterising alterations in the presence and types of proteins in target tissues and cells, proteomics holds the promise of increasing our understanding of the mechanisms of disease and identification of targets for the development of novel drug therapies. Although there are limited proteomics-based studies to date in the field of

Key words: proteomics, diabetes, two-dimensional gel electrophoresis (2DE), mass spectrometry, MALDI, ESI, SELDI. Introduction Despite continued advances in our understanding of the complex molecular mechanisms underlying the development of diabetes mellitus (DM) and its complications, it is clear that our knowledge is still limited (for reviews, see1-3). Given the predicted explosion in the number of cases of type 2 diabetes mellitus (T2DM) worldwide,4 continued research is essential, particularly with a view to understanding the impact of environmental stimuli on the physiological processes contributing to the development and progression of disease. DM and its complications arise as a consequence of defects in a variety of different tissues.3 The liver, pancreas and skeletal muscle are intimately involved in glucose homoeostasis and insulin resistance, and as such are important targets for research into the pathophysiology of diabetes and drug development. Adipose tissue is increasingly being viewed as key in the development of insulin resistance, by virtue of the activity of the hormones and cytokines it produces, and as a result of the epidemiological link between obesity and the development of T2DM/insulin resistance.5,6

Academic Unit of Molecular Vascular Medicine, The LIGHT Laboratories, Clarendon Way, University of Leeds, Leeds, LS2 9JT, UK. Eleanor Scott, Senior Lecturer in Medicine and Honorary Consultant Physician Angela Carter, Principal Research Fellow in Molecular Vascular Medicine Department of Biochemistry, Leeds University, Leeds, UK. John Findlay, Professor of Biochemistry Correspondence to: Dr Eleanor Scott Academic Unit of Molecular Vascular Medicine, The LIGHT Laboratories, Clarendon Way, University of Leeds, Leeds, LS2 9JT, UK. Tel: +44 (0)113 343 7719; Fax: +44 (0)113 343 7738 E-mail: e.m.scott@leeds.ac.uk

54
Downloaded from dvr.sagepub.com at MAHATMA GANDHI INST MED SCI on April 4, 2013

DIABETES AND VASCULAR DISEASE RESEARCH

REVIEW

Figure 1. One gene can code for many proteins as a result of splicing and post-translational modifications. This has effects on the proteins function and alters the phenotype. Gene-environment interactions also alter protein behaviour. Thus a proteome cannot be predicted from the genome
Genotype Phenotype

Figure 2. Principles of proteomics

Complex protein mixture

Splice variants

Protein or peptide separation LC

Protein separation 2DE

Protein separation SELDI chip

DNA

mRNA

Protein

Post-translational modifications
Stained gel spots imaged and analysed

Spots excised and digested

Protein exists in several forms, each with different functions

Mass spectrometry ESI-MS/MS MALDI-MS/MS

Mass spectrometry MALDI-MS

Mass spectrometry SELDI-MS

diabetes, those that have been performed offer an insight into the power and impact that proteomics is likely to have in the future, as outlined in the following sections.
Database search

Database search Protein identification

Protein profile analysis Biomarker discovery

Proteomic technologies Proteomic analyses have two different aspects expression proteomics and functional proteomics.9 Expression proteomics broadly involves the identification and quantification of proteins and characterisation of splice variants, PTMs and cellular localisation. Expression proteomics technologies involve one or more separation steps followed by identification using mass spectrometry (figure 2). The classical proteomic technique is two-dimensional gel electrophoresis (2DE), followed by protein identification by matrix-assisted laser desorption ionisation (MALDI) mass spectrometry (MS). More recent developments include liquid chromatography coupled to electrospray ionisation (ESI) MS/MS, and surfaceenhanced laser desorption ionisation (SELDI) MS. The above techniques are generally used for comprehensive protein profiling. However, they can be preceded by a variety of procedures for reduction of sample complexity, including multidimensional liquid chromatography, cellular fraction, immunoprecipitation (IP) and affinity chromatography, to name but a few. Functional proteomics includes techniques for analysis of protein-protein interactions and protein networks such as the two-hybrid systems, surface plasmon resonance (SpR) and nuclear magnetic resonance (NMR).10,11 It is beyond the scope of this review to cover all aspects of proteomics, and therefore it will focus on providing a brief overview of mass spectrometry-associated technologies and their application to the study of T2DM, to be outlined in the following sections. For detailed descriptions of the various proteomic technologies, see references.8-13 Mass spectrometry-associated proteomic analyses 2DE and MALDI-MS In this system, complex mixtures of proteins are separated by gel electrophoresis. In the first dimension, separation is based on relative charge according to isoelectric point (isoelectric

Protein identification

Database search Protein identification

focusing [IEF]); in the second dimension, separation is based on molecular mass.14 Once separated, the proteins are visualised as an array of apparent spots by treating the gel with one of a variety of protein stains, including Coomasie Blue, silver and Sypro Ruby, or by pre-staining with Cy dyes (100s to 1,000s of spots can be resolved on a single gel). The stained protein spots are imaged by specialised equipment and differences in protein expression between samples can be determined using specialised software.15 A variety of analysis programs are available which allow for comparison of multiple samples by aligning gel images. Assessment can then be made of changed protein patterns (often characteristic of PTMs) or levels (indicative of variant expression). For protein identification, protein spots of interest are excised from the gel, fragmented by proteases (most often trypsin) and the resulting mixture of peptides is then spotted onto a MALDI-MS plate. The samples are then dried and coated with MALDI matrix to promote peptide ionisation for MS analysis. This involves separation of peptides according to their time of flight (TOF), which is dependent upon mass to charge (m/z) ratio. The resultant peptide m/z fingerprint is then compared against databases (such as Mascot, www.matrixscience.com) of theoretical peptide fingerprints of all proteins to identify the protein(s) of interest.16 Liquid chromatography coupled to MS (LC-MS) In its simplest form, LC-MS involves the proteolytic digestion (most usually with trypsin) of complex mixtures of proteins,

VOLUME 2 ISSUE 2 . MAY 2005

Downloaded from dvr.sagepub.com at MAHATMA GANDHI INST MED SCI on April 4, 2013

55

REVIEW

which are then simplified by one or more liquid chromatography steps based on different peptide properties such as charge, hydrophobicity or pH. A combination of strong cation exchange chromatography followed by reversed phase chromatography is commonly used; see Wang and Hanash for a review of the variety of multidimensional chromatography approaches used in proteomics.9 Peptide analysis is (most usually) by ESI tandem MS (MS/MS). In MS/MS analysis the peptides are further resolved in MS mode: selected peptides are isolated and subjected to further fragmentation in a gas-containing collision cell before being introduced into a second phase of MS analysis to provide sequence information.10 In tandem MS analysis, different combinations of mass analyser can be coupled, including TOF, ion trap, quadrupole and Fourier transform ion cyclotron.10 ESI is most commonly coupled to ion trap and/or quadrupole mass analysers. NanoLC systems are a recent development which enable spotting onto MALDI plates for analysis, and tandem MALDI systems (either TOF/TOF or quadrupole/TOF) are increasingly being utilised for peptide identification.10 With MS/MS analysis, protein identification is more certain as linear sequences of six or more residues can be enough to identify the protein unambiguously. Often sequences can be obtained from more than one peptide.17,18 SELDI-MS In SELDI analysis the sample to be analysed is applied to a protein chip of defined binding property (e.g. hydrophobic, cation, anion, metal). This on-chip fractionation allows for retention of proteins of potential interest and reduction in sample complexity. After washing the chip to remove unwanted proteins, the energy-adsorbing matrix is added prior to laser-induced ionisation of bound proteins and TOF separation to determine m/z. For a review of SELDI technologies, see Tang et al.13 SELDI has most commonly been applied to clinical proteomics as a means of identifying protein profiles which differentiate different clinical groups, and in its simplest form provides no protein identification. SELDI analysis relies on complex bioinformatics and statistical tools to mine the protein m/z data to define the protein profiles that are unique to a particular patient group, particularly if unique markers are present in low quantities. Further downstream analyses are required, including protein digestion and peptide mass fingerprinting or MS/MS analysis, to obtain protein identification to inform further functional analyses.13 Advantages and limitations of these techniques 2DE Two-dimensional gel electrophoresis and protein analysis by spot-picking and trypsin digestion followed by MALDI-MS peptide analysis provides a number of advantages over other methods. 2DE is the best validated method for analysis of whole proteins; there is a variety of stains available, which can be utilised to obtain quantitative data using a variety of imaging and software packages for gel analysis. 2DE has the advantage of providing visualisation of the proteins and consequently, at present, it is the best method for identification

of proteins which have a variety of splice variants and posttranslational modifications. However, 2DE also has a number of limitations. Membrane proteins and extremely acidic, basic or hydrophobic proteins do not seem to be readily resolved by 2DE (in these cases 1D PAGE is often utilised, although this will not provide any information regarding post-translational modifications). In addition, 2DE does not have sufficient resolution for examining the whole dynamic range of proteome coverage simultaneously, as the currently available stains enable detection of proteins in the low ng range and above.10,11,15 This is particularly disadvantageous where the amount of sample available for analysis is limited. For abundant samples (such as plasma), a variety of fractionation methods can be used prior to 2DE to increase proteome coverage. For instance, IEF fractionation with preselection of narrow IEF ranges, or 2D liquid chromatography in conjunction with the use of large 2D gel formats (to increase sample loading), enable increased proteome coverage by improving the detection of low abundance proteins.10,11 LC-MS/MS LC-MS/MS allows for high-throughput applications, with identification of many hundreds of proteins in 24 hours, and it can cover a wider dynamic range of protein concentration than 2DE. Limitations of this method include the fact that the MS analysis does not give complete peptide coverage of a protein. Consequently, it does not readily lend itself to identification of splice variants and PTMs in complex mixtures. There are also issues regarding quantitation with this approach, although new methods for relative quantitation have been established, most notably with the introduction of staple isotope tagging of proteins, including isotope coded affinity tags (ICAT, Applied Biosystems, www.appliedbiosystems.com). ICATs comprise a biotin tag with either a light or heavy (deuterium or 13C) linker chain and a reactive group, which covalently binds to cysteine residues. Labelling two samples to be compared with the two forms of ICAT and mixing the samples prior to LC-MS/MS enables relative quantification of peptides by comparing the relative abundance of light and heavy ICAT-tagged peptides. The tag enables separation of biotin-labelled peptides prior to LCMS/MS analysis in order to reduce sample complexity. The disadvantage of this system is that it only allows for pairwise comparisons and is selective for cysteine-containing peptides. Also, the presence of the ICAT tag can impair ionisation, resulting in reduced sensitivity. A recent development which can compare up to four different samples directly is the introduction of iTRAQ (Applied Biosystems, www.appliedbiosystems. com), stable isobaric reagents which are amine-specific and therefore have the advantage that they label all peptides in a sample. The iTRAQ reagents comprise a peptide-reactive group which covalently binds to lysine and N terminal moieties of peptides, a variably sized reporter group and a balance group (which maintains equal peptide mass in MS mode). The advantage of this system over the ICAT system is that

56
Downloaded from dvr.sagepub.com at MAHATMA GANDHI INST MED SCI on April 4, 2013

DIABETES AND VASCULAR DISEASE RESEARCH

REVIEW

in MS mode peptides labelled with the different iTRAQ reagents will be observed as a single peak whereas in MS/MS mode they are readily ionised, the balance group is lost and the reporter groups are detected in the low m/z region between 113 and 119; comparison of the reporter group intensities provides relative quantification of the different labelled samples. Despite these advances, the technology is not sufficiently developed at present to enable high-throughput analysis of large clinical case control studies using this tagged protein approach. SELDI There has been much discussion of late of the relative advantages and disadvantages of SELDI-based analyses. It is clear that whilst the SELDI approach holds much promise, major issues, including reproducibility of results both within and between laboratories, need to be resolved if it is to become widely accepted as a tool in clinical proteomics. In addition, there is still the need to develop stringent and validated bioinformatics and statistical tools for mining SELDI data for predictive protein m/z profiles.19 The use of SELDI analysis solely to identify protein profiles as biomarkers for disease without subsequent identification of the protein biomarkers should also be questioned. For clinical research it will be important to determine protein biomarker identities to enable functional proteomic analyses to be carried out, which enhance both our knowledge of disease processes and the potential for therapeutic intervention. Other issues Regardless of the approach chosen for proteomic analysis, stringent procedures for the taking of samples and their processing (including storage where relevant) prior to proteomic analysis must be followed in order to ensure high-quality, reliable results. The quality of the results will be a reflection of the quality of the samples analysed. The Human Proteome Organisation (www.HUPO.org) has been established to address many of the issues related to standardisation of proteomic analysis.20 However, it is beyond the scope of this review to cover these issues. The application of proteomics to diabetes research As outlined in the introduction, there are key organs and tissues known to contribute to the development of DM and its complications. To define the underlying pathophysiological processes which contribute to disease development and to identify key pathways to target in the development of novel therapeutics will require a variety of approaches. These will include, for example, analysing tissue- and cell-specific protein profiles under conditions which mimic DM and different environmental stimuli thought to contribute to the development of DM; and analysing samples from patients with DM, with insulin resistance or impaired glucose tolerance and comparing them to healthy subjects. Outlined below are some of the key tissues and organs involved in DM and how the few proteomic analyses to date have been applied in investigating these areas.

Pancreas Insulin is synthesised in and secreted from the cells of the islets of Langerhans in the pancreas.3 The main stimulator of insulin release is glucose, and once insulin is released it exerts its effects by binding to the insulin receptor that is expressed on target cell membranes. Insulin receptor-ligand interactions stimulate tyrosine kinase activity, which is essential for insulins action. The post-receptor signalling mechanisms for insulin ultimately result in glucose being carried into cells across the cell membrane by glucose transporter proteins (GLUTs).2 Type 1 diabetes mellitus (T1DM) is characterised by a pancreatic insulin secretion defect21,22 whereas T2DM is characterised by insulin resistance in target tissues, often combined with a defect in pancreatic insulin secretion.23 Despite the fact that the pancreas is a key target organ, few proteomics studies have been performed and, to date, none have been done in human cells in relation to diabetes. The human pancreas has been explored taking a proteomics approach and a reference map of 302 proteins has been identified.24 The same has been done in mouse islets,25 and both of these will provide a useful reference for future proteomics studies of the pancreas in diabetes. Cytokines released from macrophages are known to be toxic to islet cells, and this is thought to be one of the mechanisms involved in the development of type 1 DM in genetically predisposed individuals.21 On this basis proteomic studies have been performed on rat islet cells,26 exposing them to the cytokine interleukin (IL)-1. More than 2,000 proteins were identified, of which 105 had altered expression following exposure to IL-1. The functional relevance of these changes to human disease is currently unclear. Proteomic analysis of murine pancreatic islet cells has identified proteins that are potentially involved in insulin resistance in mice since Lep/Lep mice were found to have altered actin-binding proteins, which may contribute to islet cell dysfunction and reduced insulin secretion.27 Using a proteomics approach, the effect of treatment has been studied. Peroxisome proliferator-activated receptors (PPARs) are transcription factors which through gene regulation have key roles in glucose and lipid homoeostasis. PPAR agonists (thiazolidinediones) are increasingly used in clinical diabetic practice to improve insulin sensitivity. Treatment with thiazolidinediones in Lep/Lep mice leads to increased expression of carboxypeptidase B, a protein involved in insulin processing, which suggests that PPAR agonists may normalise glucose homoeostasis in part through improved insulin processing.27 Liver The liver is pivotal in the regulation of fatty acid, lipoprotein and carbohydrate metabolism, playing a key role in glucose homoeostasis.3,28 In the fasted state the liver maintains a supply of glucose by gluconeogenesis and glycogenolysis. In the fed state insulin switches this off by inhibiting hepatic glucose production and stimulating glycogen synthesis. In insulin resistance, there is a failure to suppress hepatic glucose production and glycogenolysis. This is accompanied by fat accumulation within the hepatocytes, with increased free

VOLUME 2 ISSUE 2 . MAY 2005

Downloaded from dvr.sagepub.com at MAHATMA GANDHI INST MED SCI on April 4, 2013

57

REVIEW

fatty acid uptake and synthesis of triglycerides, leading to worsening of insulin resistance. As discussed above, PPARs are nuclear receptors involved in regulating lipid metabolism and glucose homoeostasis. PPAR is preferentially expressed in the liver and its activation leads to increased oxidation of fatty acids.29 It is therefore a target for the treatment of dyslipidaemia, and PPAR agonists normalise elevated plasma triglycerides and hyperglycaemia in insulin-resistant mice.29 Utilising a proteomics approach, the effect of a PPAR agonist on the liver proteins in insulin-resistant mice has been investigated. It was shown to upregulate at least 16 protein spots: when these were identified by MS, 14 were found to be components of peroxisomal fatty acid metabolism, confirming that PPAR induces fatty acid -oxidation and upregulates enzymes involved in lipogenesis.29,30 Muscle Skeletal muscle is the main site of glucose uptake and storage by insulin. Reduced oxidative enzyme capacity and dysfunction in skeletal muscle mitochondria, with impaired phosphorylation in insulin-resistant skeletal muscle, suggests that defects in AMP-activated protein kinase pathways are involved in the development of insulin resistance.31 There are many potential enzymes and pathways involved and consequently this field is difficult to study.2 Proteomics offers the techniques to evaluate many of these proteins simultaneously. By using the 2DE approach, skeletal muscle biopsies from 10 patients with T2DM and 10 controls were compared in one analysis.32 Eight proteins were overexpressed in T2DM skeletal muscle compared to controls. They were all found to be linked to increased cellular stress and alterations in skeletal muscle mitochondrial metabolism. In addition, two proteins that have a crucial role in ATP synthesis (CKB and ATP synthase) were underexpressed in the T2DM samples. At this stage it is not clear whether these are primary defects or changes resulting from T2DM. Future studies using healthy first-degree relatives of patients with T2DM may help to differentiate between primary defects and secondary effects. Adipose tissue Weight maintenance depends on balancing energy input with energy expenditure, which is coordinated by signals between the brain (hypothalamus/arcuate nucleus) and adipose tissue. Obesity arises when there is an imbalance between energy input and energy expenditure, resulting in adipocyte hypertrophy (increased cell size) and hyperplasia (increased cell number). The consequences of obesity include increased secretion of adipocyte proteins, including inflammatory mediators, angiotensinogen and plasminogen activator inhibitor-1 (PAI-1), and these proteins have been associated with the development of T2DM.33 Obesity is rapidly becoming one of the major health care issues of the 21st century. This is mirrored by increases in the pathological consequences of obesity, with classical T2DM being diagnosed in children and adolescents for the first time. There is growing evidence that the adipocytes of obese

individuals produce and secrete proteins which promote the development of T2DM.34,35 It is clear that abnormal adipocyte function is associated with T2DM, although it remains unclear what the underlying mechanisms promoting the onset of T2DM are and why some obese individuals do not go on to develop complications such as T2DM. Novel adipocyte proteins (adiponectin, for example) continue to be identified and characterised, and consequently an understanding of both the cellular and secreted adipocyte protein complement and the factors associated with pre-adipocyte differentiation will provide insights into the mechanisms involved in the regulation of adiposity and the pathogenesis of obesity and T2DM. It has been suggested that adipocytes have a finite capacity for lipid storage (and consequently a finite size) and that fully lipid-laden adipocytes secrete factors to promote pre-adipocyte proliferation and differentiation into mature adipocytes in order to facilitate further lipid storage. Some individuals appear to have a defect in this differentiation process, leading to increased hypertrophic adipocytes which have been reported to secrete excess proteins involved in inflammatory processes (e.g. tumour necrosis factor [TNF]), consequently contributing to the pathological consequences of obesity. Limited proteomic analyses have been carried out in the murine 3T3 L1 cell line to determine variations in cellular and secreted proteins during differentiation.36-39 An increase in a variety of mitochondrial proteins, which was associated with observed differences in mitochondrial morphology, was observed during differentiation and treatment with the PPAR agonist rosiglitazone.37 Analyses of membrane and secreted proteins during differentiation identified a number of proteins not previously known to be secreted by adipocytes38,39 and membrane proteins selectively expressed by adipocytes.36 Analysis of the cellular and secreted proteins of preadipocytes and adipocytes from normoglycaemic individuals and from individuals with insulin resistance, impaired glucose tolerance and T2DM to determine differences in quantity and structural modifications (splice variants, glycation etc.) will provide invaluable insights into the mechanisms underlying the pathogenesis of T2DM. Heart Diabetic cardiomyopathy is a recognised complication of DM, leading to congestive cardiac failure. Proteomics has been applied to the study of cardiac proteins in a type 1 DM mouse model.40 Twenty proteins were altered in mice with cardiomyopathy compared to controls, 12 of them mitochondrial proteins. This may reflect oxidative stress damaging the mitochondria in DM and contributing to the development of cardiomyopathy. Kidney and urine Diabetic nephropathy is the most common cause of renal failure worldwide.41 Using a mouse model of type 1 DM nephropathy, the renal proteome was compared to that of non-diabetic mice.42 Whole kidneys were lysed and proteins separated by 2DPAGE, then identified by MALDI-TOF. Of those proteins that were altered, it was noted that elastase

58
Downloaded from dvr.sagepub.com at MAHATMA GANDHI INST MED SCI on April 4, 2013

DIABETES AND VASCULAR DISEASE RESEARCH

REVIEW

IIIB was reduced and accompanied by an increase in monocyte neutrophil elastase inhibitor, with accumulation of elastin throughout the kidney. Subsequent examination of human renal biopsies by immunohistochemistry of elastin confirmed increased elastin expression, suggesting it has a role in the pathophysiology of diabetic nephropathy. It is widely accepted that early identification of patients who are at risk of developing diabetic nephropathy is important, and urine is currently screened for the presence of microalbumin.41 Proteomic techniques have thus been applied to examine the urine of healthy subjects and to compare the proteome seen with that of the proteome of the urine of diabetic subjects with and without albuminuria.43 Diabetic subjects without albuminuria had noticeably different polypeptide patterns to controls, and in those subjects with albuminuria a different polypeptide profile again was seen. Using MS/MS, three of the discriminating polypeptides have been identified, and they are small fragments of proteins known to be affected by the presence of renal disease. Because of their small size it would seem possible that they could serve as very early indicators of renal damage, before the glomerular filter has become damaged enough to let through the albumin that we currently use to identify renal disease. Plasma and serum Extracellular fluids contain many secreted proteins and these change in response to the state of the organism. With disease it is likely that there may be changes in the protein composition of extracellular fluids. In comparison to the tissues outlined above, plasma (and serum) also has the advantage of being easy to obtain in relatively large quantities, which is an obvious advantage for clinical research. Patients are also more likely to provide consent for a blood sample in preference to tissue biopsies. Plasma is a valuable resource for proteomic analysis, since it comprises components of virtually every physiological and pathological process occurring in the human body. Since T2DM is associated with defects in multiple organs, proteomic analysis of plasma can provide much information relevant to the study of diabetes and its complications and has the potential to provide diagnostic markers. As Anderson and Anderson describe: The human plasma proteome holds the promise of a revolution in disease diagnosis and therapeutic monitoring provided that major challenges in proteomics and related disciplines can be addressed. Plasma is not only the primary clinical specimen but also represents the largest and deepest version of the human proteome present in any sample.44 It is apparent that at present we have identified only a minor fraction of the total protein variants present in plasma, estimated to be in the order of 500,000 (excluding the immunoglobulin fraction) if the different splice variants and PTMs are included.44 The main challenge in the proteomic analysis of plasma is the wide dynamic range of protein concentration (in the region of 10 orders of magnitude), which has implications for the analysis of lower concentration components.44 Initiatives such as the Human Proteome Organisation

(HUPO) Plasma Proteome Project (PPP) are enabling the standardisation of plasma proteomic analyses in the clinical context.45 Furthermore, improved systems for easy sample fractionation, image analysis and data analysis, and methods for depletion of high abundance proteins, have made possible the study of plasma samples from clinical cohorts. For most applications the first step in the proteomic analysis of plasma (or serum) involves depletion of the most abundant proteins (in particular albumin and immunoglobulin) by a variety of methods to enable analysis of the less abundant proteins. Additional fractionation steps will also increase plasma proteome coverage. To date, only one proteomic analysis of serum has been reported in relation to DM.46 However, as no clinical characterisation of the study or control group was made it is difficult to determine and interpret the relevance of their findings. Proteomics of serum, plasma or even peripheral blood cells has potential utility for closer understanding of disease mechanisms, detecting disease markers, new targets for drug treatment, and the mechanisms involved in drug action. It could help to develop new markers for the presence of disease and may eventually allow more effective screening of diabetic subjects who are at risk of diabetic complications. Conclusion Proteomic technologies will enable the analysis and comparison of protein expression profiles in tissues and plasma or serum from people with diabetes and normal controls, in animal models of disease and in response to drug treatment. The concept of proteomics is straightforward but putting it into practice remains challenging due in part to the complex dynamic nature of the proteome and the seemingly limitless data that could be generated. It is also expensive and timeconsuming, and should not be undertaken lightly. If applied appropriately, however, proteomics offers the ability to identify proteins involved in the development of insulin resistance, dysglycaemia and DM, and an understanding of how environmental factors can influence the development and progression of disease. It has the potential to identify protein markers of susceptibility to micro- and macrovascular diseases and to help our understanding of the mechanisms of drug action, whilst identifying new targets for therapeutic development. Conflict of interest None declared.
1. Kelly MA, Rayner ML, Mijovic CH, Barnett AH. Molecular aspects of type 1 diabetes. Mol Pathol 2003;56:1-10. 2. White MF. IRS proteins and the common path to diabetes. Am J Physiol Endocrinol Metab 2002;283:E413-E422. 3. Pickup J, Williams G. Textbook of Diabetes. Oxford: Blackwell Science, 1997. 4. Zimmet P . The burden of type 2 diabetes: are we doing enough? Diabetes Metab 2003;29:6S9-18. 5. Hsueh WA, Law R. The central role of fat and effect of peroxisome proliferator-activated receptor-gamma on progression of insulin resistance and cardiovascular disease. Am J Cardiol 2003;92:3J-9J. 6. Goldfine AB, Kahn CR. Adiponectin: linking the fat cell to insulin sensitivity. Lancet 2003;362:1431-2. 7. Stern MP . Do non-insulin-dependent diabetes mellitus and cardiovas-

References

VOLUME 2 ISSUE 2 . MAY 2005

Downloaded from dvr.sagepub.com at MAHATMA GANDHI INST MED SCI on April 4, 2013

59

REVIEW

8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26.

27. 28. 29.

cular disease share common antecedents? Ann Intern Med 1996;124: 110-16. Banks RE, Dunn MJ, Hochstrasser DF et al. Proteomics: new perspectives, new biomedical opportunities. Lancet 2000;356:1749-56. Wang H, Hanash S. Multi-dimensional liquid phase based separations in proteomics. J Chromatogr B Analyt Technol Biomed Life Sci 2003; 787:11-18. Aebersold R, Mann M. Mass spectrometry-based proteomics. Nature 2003;422:198-207. Zhu H, Bilgin M, Snyder M. Proteomics. Annu Rev Biochem 2003;72: 783-812. Tyers M, Mann M. From genomics to proteomics. Nature 2003;422: 193-7. Tang N, Tornatore P , Weinberger SR. Current developments in SELDI affinity technology. Mass Spectrom Rev 2004;23:34-44. O'Farrell PH. High resolution two-dimensional electrophoresis of proteins. J Biol Chem 1975;250:4007-21. Sivakumar A. 2D gels and bioinformatics an eye to the future. In Silico Biol 2002;2:507-10. Henzel WJ, Watanabe C, Stults JT. Protein identification: the origins of peptide mass fingerprinting. J Am Soc Mass Spectrom 2003;14:93142. Mann M, Wilm M. Error-tolerant identification of peptides in sequence databases by peptide sequence tags. Anal Chem 1994;66:4390-9. Perkins DN, Pappin DJ, Creasy DM, Cottrell JS. Probability-based protein identification by searching sequence databases using mass spectrometry data. Electrophoresis 1999;20:3551-67. Baggerly KA, Morris JS, Coombes KR. Reproducibility of SELDI-TOF protein patterns in serum: comparing datasets from different experiments. Bioinformatics 2004;20:777-85. Hanash S. Building a foundation for the human proteome: the role of the Human Proteome Organization. J Proteome Res 2004;3:197-9. Nerup J, Mandrup-Poulsen T, Helqvist S et al. On the pathogenesis of IDDM. Diabetologia 1994;37(suppl 2):S82-S89. Nerup J, Mandrup-Poulsen T, Molvig J, Helqvist S, Wogensen L, Egeberg J. Mechanisms of pancreatic beta-cell destruction in type I diabetes. Diabetes Care 1988;11(suppl 1):16-23. Chiasson JL, Rabasa-Lhoret R. Prevention of type 2 diabetes: insulin resistance and beta-cell function. Diabetes 2004;53(suppl 3):S34-S38. Hu L, Evers S, Lu ZH, Shen Y, Chen J. Two-dimensional protein database of human pancreas. Electrophoresis 2004;25:512-18. Sanchez JC, Chiappe D, Converset V et al. The mouse SWISS-2D PAGE database: a tool for proteomics study of diabetes and obesity. Proteomics 2001;1:136-63. Sparre T, Christensen UB, Gotfredsen CF et al. Changes in expression of IL-1 beta influenced proteins in transplanted islets during development of diabetes in diabetes-prone BB rats. Diabetologia 2004;47: 892-908. Sanchez JC, Converset V, Nolan A et al. Effect of rosiglitazone on the differential expression of diabetes-associated proteins in pancreatic islets of C57Bl/6 lep/lep mice. Mol Cell Proteomics 2002;1:509-16. Collier JJ, Scott DK. Sweet changes: glucose homeostasis can be altered by manipulating genes controlling hepatic glucose metabolism. Mol Endocrinol 2004;18:1051-63. Edvardsson U, Alexandersson M, Brockenhuus VL et al. A proteome analysis of livers from obese (ob/ob) mice treated with the peroxisome

proliferator WY14,643. Electrophoresis 1999;20:935-42. 30. Edvardsson U, von Lowenhielm HB, Panfilov O, Nystrom AC, Nilsson F, Dahllof B. Hepatic protein expression of lean mice and obese diabetic mice treated with peroxisome proliferator-activated receptor activators. Proteomics 2003;3:468-78. 31. Winder WW, Hardie DG. AMP-activated protein kinase, a metabolic master switch: possible roles in type 2 diabetes. Am J Physiol 1999; 277:E1-E10. 32. Hojlund K, Wrzesinski K, Larsen PM et al. Proteome analysis reveals phosphorylation of ATP synthase beta-subunit in human skeletal muscle and proteins with potential roles in type 2 diabetes. J Biol Chem 2003;278:10436-42. 33. Fruhbeck G, Gomez-Ambrosi J, Muruzabal FJ, Burrell MA. The adipocyte: a model for integration of endocrine and metabolic signaling in energy metabolism regulation. Am J Physiol Endocrinol Metab 2001;280:E827-E847. 34. Considine RV, Nyce MR, Morales LM et al. Paracrine stimulation of preadipocyte-enriched cell cultures by mature adipocytes. Am J Physiol 1996;270:E895-E899. 35. Gabriely I, Ma XH, Yang XM et al. Removal of visceral fat prevents insulin resistance and glucose intolerance of aging: an adipokine-mediated process? Diabetes 2002;51:2951-8. 36. Xu A, Choi KL, Wang Y et al. Identification of novel putative membrane proteins selectively expressed during adipose conversion of 3T3-L1 cells. Biochem Biophys Res Commun 2002;293:1161-7. 37. Wilson-Fritch L, Burkart A, Bell G et al. Mitochondrial biogenesis and remodeling during adipogenesis and in response to the insulin sensitizer rosiglitazone. Mol Cell Biol 2003;23:1085-94. 38. Choi KL, Wang Y, Tse CA, Lam KS, Cooper GJ, Xu A. Proteomic analysis of adipocyte differentiation: Evidence that alpha2 macroglobulin is involved in the adipose conversion of 3T3 L1 preadipocytes. Proteomics 2004;4:1840-8. 39. Kratchmarova I, Kalume DE, Blagoev B et al. A proteomic approach for identification of secreted proteins during the differentiation of 3T3-L1 preadipocytes to adipocytes. Mol Cell Proteomics 2002;1:213-22. 40. Shen X, Zheng S, Thongboonkerd V et al. Cardiac mitochondrial damage and biogenesis in a chronic model of type 1 diabetes. Am J Physiol Endocrinol Metab 2004;287:E896-E905. 41. Marshall SM. Recent advances in diabetic nephropathy. Clin Med 2004;4:277-82. 42. Thongboonkerd V, Barati MT, McLeish KR et al. Alterations in the renal elastin-elastase system in type 1 diabetic nephropathy identified by proteomic analysis. J Am Soc Nephrol 2004;15:650-62. 43. Mischak H, Kaiser T, Walden M et al. Proteomic analysis for the assessment of diabetic renal damage in humans. Clin Sci (Lond) 2004;107: 485-95. 44. Anderson NL, Anderson NG. The human plasma proteome: history, character, and diagnostic prospects. Mol Cell Proteomics 2002;1:84567. 45. Omenn GS. The Human Proteome Organization Plasma Proteome Project pilot phase: reference specimens, technology platform comparisons, and standardized data submissions and analyses. Proteomics 2004;4:1235-40. 46. Zhang R, Barker L, Pinchev D et al. Mining biomarkers in human sera using proteomic tools. Proteomics 2004;4:244-56.

60
Downloaded from dvr.sagepub.com at MAHATMA GANDHI INST MED SCI on April 4, 2013

DIABETES AND VASCULAR DISEASE RESEARCH