Proteomics as a Tool for Discovery: Proteins Implicated in Alzheimer’s Disease are Highly Expressed in Normal Pancreatic Islets Mark R. Nicolls*,,†,‡ Jason M. D’Antonio,§ John C. Hutton,‡ Ronald G. Gill,‡ Jennifer L. Czwornog,† and Mark W. Duncan§ Division of Pulmonary Sciences and Critical Care Medicine, University of Colorado Health Sciences Center, Barbara Davis Center for Childhood Diabetes, University of Colorado Health Sciences Center, and Biochemical Mass Spectrometry Facility, School of Pharmacy, University of Colorado Health Sciences Center, Denver, Colorado 80262 Received October 3, 2002
A proteomic analysis of islets was undertaken to determine the protein constituents of normal adult mouse islets. Unexpectedly, we identified several islet proteins that are associated with the pathogenesis of Alzheimer’s disease. Some of these proteins had chaperone activity that is integral to proper protein folding. This group includes GRP78, valosin-containing protein, calreticulin, protein disulfide isomerase, DnaK, HSP70, HSP60, and TCP-1. Additionally, neuronal proteins key to coordinated neuronal guidance and survival were also identified in islets. This group includes proprotein convertase subtilisin, collapsin response mediator protein 2, ubiquinol-cytochrome c reductase core protein, L-3-hydroxyacyl-Coenzyme A dehydrogenase, glutamine synthetase, peroxiredoxin, and secretogogin. An important subset of the proteins identified here has not been reported previously in pancreatic islets. Abnormal activity of these proteins in brain may contribute to the pathogenesis of Alzheimer’s disease, a neurodegenerative condition characterized by focal amyloid deposits with neurofibrillary tangles. The putative role of these proteins in Alzheimer’s pathogenesis is intriguing given the possible clinical relationship and pathological similarity of Alzheimer’s disease to type 2 diabetes. These findings have therefore led to the hypothesis that these proteins may also play a role in type 2 diabetes. Keywords: proteomics • islets • type 2 diabetes • Alzheimer’s disease • chaperone
1. Introduction Proteomics has considerable promise to improve our understanding of health and disease, but to realize its full potential, studies must be well designed, data must be carefully validated, and the findings must be rationalized and interpreted in the context of our current knowledge. One of the great strengths of proteomics is its potential to aid in the generation of novel hypotheses and/or to lead to new insights that may not have been considered previously. Our laboratory has an interest in protein expression in islet transplantation for the treatment of type 1 diabetes with the objective of gaining a better understanding of allograft tolerance and rejection. In particular, we are concerned with specific and early biomarkers of both tolerance and rejection, and with defining parameters that predict optimal islet viability and * To whom correspondence should be addressed. Mark R. Nicolls, MD, Box B140, Barbara Davis Center for Childhood Diabetes, 4200 E. 9th Ave, Denver, CO 80262. Telephone: (303) 315-5056. Fax: (303) 315-4124. E-mail:
[email protected]. † Division of Pulmonary Sciences and Critical Care Medicine University of Colorado Health Sciences Center. ‡ University of Colorado Health Sciences Center. § Biochemical Mass Spectrometry Facility School of Pharmacy, University of Colorado Health Sciences Center. 10.1021/pr025576x CCC: $25.00
2003 American Chemical Society
function. As a first step in reaching this objective we have begun defining the protein profile of the normal mouse islet. Our approach is based on 2-dimensional (2D) gel electrophoresis and matrix-assisted laser desorption ionization time-of-flight (MALDI-TOF) mass spectrometry (MS) and is complementary to functional genomics that is being used to generate a large database as a resource for diabetes research.1 This study catalogues islet proteins of the normal adult mouse islet. In the course of defining islet protein function, we identified islet proteins that have that have been implicated in the pathogenesis of Alzheimer’s disease (AD). These proteins can broadly be divided into 2 categories: (1) proteins with chaperone activity (i.e., critical for normal protein processing) and (2) proteins associated with neuronal function. These proteins include GRP78, valosin-containing protein, calreticulin, protein disulfide isomerase, DnaK, HSP70, HSP60, and TCP1, proprotein convertase subtilisin, collapsin response mediator protein 2, ubiquinol-cytochrome c reductase core protein, L-3hydroxyacyl-Coenzyme A dehydrogenase, glutamine synthetase, peroxiredoxin, and secretogogin. The presence of these proteins is compelling given the already established clinical associations and biological similarities between type 2 diabetes and AD. Journal of Proteome Research 2003, 2, 199-205
199
Published on Web 01/11/2003
research articles In type 2 diabetes, the mechanism for islet amyloid formation is likely multifactorial, involving multiple cellular stressors including persistent hyperglycemia, subsequent increased insulin production with islet amyloid polypeptide (IAPP, amylin) cosecretion, oxidative damage, and high dietary fat.2 The progressive accumulation of IAPP-containing deposits ultimately replaces β-cell mass with amyloid and contributes to the hyperglycemia of type 2 diabetes.3 In AD, amyloid deposits and associated neurofibrillary tangles have been implicated in the evolution of dementia in AD.4 AD amyloid develops from the polypeptide A-beta 40 (Aβ). In addition to cataloging islet proteins, this study discusses certain islet proteins with regard to their possible role in the development of AD. As a discovery and hypothesis-generating tool, the proteomic analysis of pancreatic islets is subsequently directing promising avenues of research into the pathogenesis of type 2 diabetes.
2. Materials and Methods Islet Extraction. This study conforms to the use and care guidelines of laboratory animals as set forth by the University of Colorado Institutional Animal Care and Use Committee. Islets were isolated from 3- to 6-month-old male BALB/c mouse pancreata by collagenase digestion.5 Islets were handpicked, washed three times in Hank’s Balanced Saline Solution (HBSS), and stored at -80 °C. Samples were freeze/thawed twice prior to analysis. Sample Preparation. Four-hundred fifty islets were lysed in 1 mL lysis buffer (1%Triton X-100, 1 mM Tris, 5 mM EDTA, 50 mM NaCl, 50 mM NaF, 1 mM Na3VO4 (sodium vanadate), and then centrifuged at 13 000 rpm for 30 min at 4 °C. The total protein concentration in the supernatant was determined by the Coomassie Plus-200 Bradford protein assay (Pierce Chemical Company, Rockford, IL). Protein concentrations were in the range of 0.5-1.0 µg/µL. Protein (500 µg) was then precipitated (4:1, methanol:chloroform), the sample centrifuged, the supernatant removed, and the protein pellet dissolved in 300 µL 1st dimension rehydration buffer containing 9M urea, 4% CHAPS, 65 mM DTT, 35 mM Tris, 0.8% carrier ampholytes (3-10 NL), and 0.5% bromophenol blue. Gel Electrophoresis and In-gel Digestion. Immobiline DryStrips (pH 3-10) were rehydrated with the 500 µg solubilized protein in an Immobiline DryStrip Reswelling Tray (Amersham Pharmacia Biotech, Pescataway, NJ). Isoelectric focusing was performed on a Multiphor II electrophoresis unit (Amersham Pharmacia Biotech, Pescataway, NJ). Equilibration of the first dimension strip after isoelectric focusing occurs in two steps: reduction followed by alkylation. Each step was performed for 10 min on a standard benchtop rocker. The reduction solution contained 1% DTT, 6 M urea, 2% SDS, 15% glycerol, and 0.05 M Tris pH 6.8. DTT is a reducing agent that cleaves disulfide bonds. The alkylation solution contained 6 M Urea, 2% SDS, 15% glycerol, and 0.05 M Tris pH 6.8 and 1.25% iodoacetamide that provided protection to S-H groups by alkylation. The second dimension was run on 10% polyacrylamide SDS gels (BioRad Protean II xi Cell, BioRad, Hercules, CA). The two-dimensional (2D) gels were immediately stained with G-250 Coomassie Blue and digitally imaged (UMAX Powerlook III, UMAX Data Systems, Inc., Taiwan). Individual spots were excised manually into a microtube and immediately covered in 25 mM ammonium bicarbonate. Spots were temporarily stored in this solution overnight at 4 °C, or for subsequent mass spectrometry analysis, the ammonium bicarbonate was removed after 30 min and the spots equili200
Journal of Proteome Research • Vol. 2, No. 2, 2003
Nicolls et al.
brated with a 1:1 solution of 100% CH3CN (acetonitrile): 25 mM ammonium bicarbonate. The liquid was then removed by aspiration and the gel pieces destained (serial 200 µL washes of CH3CN: 25 mM NH4HCO3, 1:1, until the G-250 Coomassie Blue stain was removed. The liquid was then aspirated and sequencing grade porcine modified trypsin was added (10 µL, 0.01 mg/mL, Promega, Madison, WI). The proteins were digested in-gel (37 °C) and stored at 4 °C until analysis.6 Aliquots (0.5 µL) were taken directly from this solution and analyzed by MALDI-TOF MS. MALDI-TOF-MS and Protein Identification. Mass analysis was performed by MALDI-TOF-MS (Voyager-DE-STR mass spectrometer, Applied Biosystems/Perseptive, Foster City, CA) in reflectron, delayed extraction mode. Samples (0.5 µL) were spotted on a standard target, overlaid with matrix (R-cyano4-hydroxycinnamic acid, 0.5 µL, 5 mg/mL), and the samplematrix mixture was dried at room temperature. Spectra were the sum of 100 laser shots and those peaks with a signal-tonoise ratio of greater than 5:1 were selected for database searching. Spectra were internally calibrated using autolytic trypsin peptides (m/z 515.33, 842.51, 1045.56, 2211.10). Individual spectra were baseline corrected, smoothed, and deisotoped. The monoisotopic masses for each protonated peptide were entered into Mascot (http://matrixscience.com), an algorithm that searches the NCBI database to generate statistically significant peptide mass fingerprint identifications. Masses derived from trypsin, matrix, keratin, and G-250 Coomassie Blue were excluded from the analysis. Search parameters included the molecular weight (15-150 kilodaltons), pI (3-10), carbamidomethylation of cysteine residues, maximum average mass error of 50 ppm, and one missed cleavage per peptide. The database search was not limited to looking for mouse proteins because information on the mouse is incomplete. By searching the nonredundant database, homology searches could also be performed. Protein identification required that all of the following criteria be met: (1) p < 0.05; (2) observed MW and pI as measured from the 2D gel within ( 20% of the predicted values; and (3) one of the following: a. for average mass errors of less than 10 ppm, five peptides matched or sequence coverage of 15%; b. for average mass errors 10-20 ppm, six peptides matched or sequence coverage of 18%; c. for average mass errors 20-30 ppm, seven peptides matched or sequence coverage of 22%; d. for average mass errors 30-50 ppm, eight peptides matched or sequence coverage of 25%; and (4) no more than 50 peaks submitted.
3. Results and Discussion A method was developed for 2D gel analysis of mouse pancreatic islets that resolved proteins between ∼15-150 kDa (Figure 1) and which when combined with MALDI-TOF mass spectrometry led to the identification of a set of mouse proteins (Table 1). (Matches against homologues from other species where not included in the list of identified proteins.) Insulin 1 and insulin 2 were identified by direct analysis of the cell lysate by MALDI-TOF-MS (i.e., without prior 2D gel analysis). The molecular weights of insulin 1 and 2 were 5796 and 5708 Da, respectively. The majority of the other proteins were noted in replicate gels (indicated by “n”). Not all proteins were identified in all gels analyzed. Up to 150 spots were excised from each
Alzheimer’s-Related Proteins in Islets
research articles
Figure 1. 2-D PAGE of adult murine islets. Gels were stained with Coomassie blue on gels with a pH range of 3.0-10.0. These results are representative of seven separate experiments.
gel. Utilizing the stringent criteria outlined in the Methods section, up to 40% of proteins excised from the gels did not meet the criteria for identification. Not all of the excised proteins could be identified for several reasons. These included the following: (a) comigration that resulted in a mixture of proteins at one location on the gel, (b) low abundant proteins that did not yield sufficient signal, (c) the presence of modifications, or (d) because the protein under question was not present in the database. This last point is relevant to mouse proteins because homology searches are not always reliable or accurate, and the complete mouse genome is not publicly available. At the time of this study, approximately 80 000 mouse proteins were present in the nonredundant database and we estimate that this represents less than 50% of all mouse proteins. When a protein was identified on only one or two gels, it was likely that the appropriate spot(s) was not excised and analyzed from the gels that were run, but we cannot rule out the possibility that these entries represent unique proteins from individual samples. Proteins were categorized by function and location (Figure 2). The majority of identified proteins were cytosolic or mitochondrial. The paucity of membrane and nuclear proteins is a consequence of the protein extraction protocol. In addition to a large number of metabolic proteins, chaperone and neurally associated proteins were also predominant. The SWISS-2D PAGE database established at the Geneva Proteomics Center has catalogued mouse islet proteins along with liver, muscle, and adipose tissue with a view to assist studies of diabetes and obesity.7 In this study, 44 C57Bl/6 islet proteins were described. Fifty percent of these proteins were also identified in the current project utilizing BALB/c islets. Proteins not common to both studies may reflect differences in islet isolation techniques, protein extraction methods, strain variation or data analysis (e.g., different database searching).
Figure 2. Distribution of functions (a) and subcellular locations (b) of proteins represented by the data set presented in Table 1.
Sanchez and co-workers at the Geneva Proteomics Center have also recently reported differential protein expression in pancreatic islets from mice prone to glucose intolerance.8 These investigators discovered 9 proteins that were differentially expressed in obese (lep/lep) mice compared to lean mice. Subsequently, four of these proteins were found to be significantly modulated in lep/lep mice following treatment with the insulin sensitizer drug, rosiglitazone. Similar studies have targeted the identification of proteins specifically affected by nitric oxide (NO) and interleukin-1β stress.9,10 The current study extends the description of the protein constituents of normal adult mouse islets. Given that thousands of spots are available for identification and that each spot represents a minimum of one unique protein, the currently identified proteome represents only a small fraction of the whole islet proteome. Of interest, some proteins identified in the current study have also been reported in AD literature. Most of these ADrelated proteins fell into two general categories: (1) proteins with molecular chaperone activity and (2) proteins that were neurally associated. We subsequently examined the entire Journal of Proteome Research • Vol. 2, No. 2, 2003 201
research articles
Nicolls et al.
Table 1. Adult Mouse Islet Proteins Identified by 2D Gel Electrophoresis and MALDI-TOF-MS protein name
accession no.
predicted M. W./pI
observed M. W.(kDa)/pI
chaperone activity
58 kDa glucose regulated protein 78 kDa glucose-regulated protein precursor (GRP 78) 60 kDa heat shock protein, mitchondrial precursor aconitase 2, mitochondrial actin, gamma, cytoplasmic adenylyl cyclase - associated CAP protein Aldehyde reductase aldo-keto reductase family 1, member C12 aldolase 1, A isoform similar to aldose reductase alpha enolase alpha glucosidase 2, alpha neutral subunit amylase 2, pancreatic annexin A4; annexin IV annexin A5 similar to archain 1 ATP citrate (pro-S) lyase similar to ATP synthase, H+ transporting mitochondrial F1 complex, beta subunit ATP synthase, H+ transporting mitochondrial F1, alpha subunit ATPase, H+ transporting, lysosomal bifunctional purine biosynthesis protein PURH calcium binding protein, 140 kDa; 170 kDa GRP precursor; 150 kDa oxygen reg. prot. calreticulin carboxyl ester lipase carboxypeptidase H precursor chaperonin subunit 2 (beta) chaperonin subunit 5 (epsilon) collapsin response mediator protein 2 similar to dihydrolipoamide dehydrogenase elongation factor 1-alpha 1 elongation factor 1-gamma elongation factor 2 similar to elongation factor Tu ERp44 esterase 10; esterase D expressed in nonmetastatic cells 2, protein (NM23B) fumarate hydratase 1 glutamate dehydrogenase similar to glutamine synthetase similar to glutathione reductase 1 glyceraldehyde-3-phosphate dehydrogenase glycerol phosphate dehydrogenase 1, mitochondrial guanine nucleotide binding protein beta 1 subunit guanine nucleotide binding protein beta 2 related sequence guanine nucleotide binding protein, beta 2 subunit guanosine diphosphate (GDP) dissociation inhibitor 3 GDI beta heat shock protein 70 kDa heat shock protein, 74 kDa heterogeneous nuclear ribonucleoprotein A2 similar to hydroxyacyl-Coenzyme A dehydrogenase isocitrate dehydrogenase 2 (NADP+), mitochondrial isovaleryl coenzyme A dehydrogenase L-3-hydroxyacyl-Coenzyme A dehydrogenase malate dehydrogenase, mitochondrial malate dehydrogenase, soluble peptidylprolyl isomerase A; cyclophilin A
20911547 2506545
57/5.9 72/5
58/5.8 75/4.9
X X
5 5
3219998
61/6
60/5.6
X
4
18079339 6752954 6671666 10946870 20889049 6671539 20916042 13637776 6679891 6753052 7304889 13277612 20888157 21263374 20867693
86/8.1 42/5.3 52/7.2 37/6.9 37/6.2 40/8.3 36/6.7 47/6.4 110/5.7 58/7.2 36/5.4 36/4.8 58/5.9 120/7.1 56/5.2
80/7.9 42/5.2 53/7.7 37/7.8 35/6.7 38/8.7 36/7.5 48/6.5 110/5.7 50/7.2 33/5.4 35/4.8 63/6.6 45/7 50/5.5
6680748
60/9.2
52/7.9
7
6680752 15214215
67/5.6 65/6.3
71/5.6 62/7
4 2
12831229
110/5.1
140/5.1
4
6680836 6753406 3287958 6671700 6671702 6753676 20847690 1169475 13626388 18202285 20843609 19072792 13937355 6679078
48/4.3 66/5.9 54/5.1 58/6 60/5.7 63/6 55/8 50/9.1 50/6.3 96/6.4 50/6.8 47/5.1 32/6.7 17/7
63/4.3 73/6.3 54/4.8 52/6.5 60/5.8 63/6.4 57/7 50/9.1 50/6.7 90/7.5 47/6.8 44/5.1 33/7.2 17/7.3
20831568 6680027 20825772 20861909 6679937
50/7.8 61/8 42/6.64 54/8.2 36/8.4
40/8.1 56/7.2 42/7.1 50/7.8 36/8.5
X
1 5 2 1 3
20827616
81/6.2
73/6.3
X
4
6680045
38/5.6
36/5.6
1
12848861
31/7.7
30/7.3
1
13937391
38/5.6
37/5.6
4
6679987
51/5.9
44/6.3
1
13624307 20877586 7949053 20824854
71/5.4 74/5.8 36/8.7 51/9.4
70/5.5 74/5.7 25/8 40/9
6680343
51/8.9
39/8.8
9789985 6680163 6678916 6678918 6679439
47/8.5 35/8.8 36/8.9 36/6.2 18/7.7
41/6.8 34/8.3 36/8.9 36/6.2 18/7.9
202
Journal of Proteome Research • Vol. 2, No. 2, 2003
neurally associated
n)
5 4 2 1 1 3 1 6 3 6 1 4 2 1 5
X
X X X X X X X X X
X X X
5 2 4 1 2 3 3 4 5 4 2 3 1 2
4 5 1 1 1
X
1 5 6 5 1
research articles
Alzheimer’s-Related Proteins in Islets Table 1 (Continued) protein name
accession no.
predicted M. W./pI
observed M. W.(kDa)/pI
peroxiredoxin 1 peroxiredoxin 5 similar to phosphoenolpyruvate carboxykinase 2 [RIKEN cDNA 9130022B02] phosphoglycerate kinase 1 phosphoglycerate mutase 1 phosphoribosyl pyrophosphate synthetase 1 similar to propionyl CoA-carboxylase alphasubunit proprotein convertase subtilisin/kexin type 1 protein disulfide isomerase precursor protein disulfide isomerase precursor A4 (Erp72) similar to protein disulfide isomerase-related protein [RIKEN cDNA 1700015E05] pyruvate decarboxylase pyruvate kinase 3 RAS-like, family 2, locus 9 similar to retinal degeneration slow protein S-adenosylhomocysteine hydrolase similar to secretagogin precursor serum albumin precursor [Bos taurus] (contaminant of islet preparation) similar to succinate dehydrogenase complex, subunit A, flavoprotein T complex protein 1 transketolase tryptophan tRNA synthetase tubulin alpha 6 tubulin, beta 5 tumor rejection antigen gp96 tyrosine 3-monooxygenase (tyrosine hydroxylase) [epsilon polypeptide] tyrosine 3-monooxygenase/tryptophan 5monooxygenase [gamma polypeptide] ubiquinol - cytochrome c reductase core protein similar to 2-oxoglutarate dehydrogenase E1 component, mitochondrial precursor valosin-containing protein voltage dependent anion channel 1 voltage dependent anion channel 2
547923 6671549 20876078
22/8.3 25/6 71/6.9
20/8.1 25/6 70/6.9
6679291 20178035 20853786 20882127
45/7.5 29/6.7 35/6.5 80/6.8
39/8.3 28/6.7 32/7.2 75/6.5
7305371 129729 119531 20845230
85/6 58/4.8 72/5.1 48/5
72/5.1 55/4.7 73/5.2 47/5.1
6679237 6755074 6677677 20900131 7709980 20345245 1351907
130/6.3 58/7.2 25/7 83/5 48/6 32/5 69/5.8
120/6.5 56/7.5 25/7.8 84/4.9 45/6.6 29/5 73/5.7
20908717
73/7.1
68/6.7
20897385 6678359 6755991 6678469 7106439 6755863 6678619
60/5.8 68/7.2 54/6.3 51/5 50/4.8 92/4.7 29/4.6
60/6.2 65/7.8 55/6.9 55/5 50/4.9 95/4.7 31/4.5
9256646
29/4.8
13384794
chaperone activity
neurally associated
X X
n)
1 2 3 1 3 2 2
X X X X
X X
2 5 2 5 5 5 1 2 3 1 4 1
X
1 4 2 3 4 4 3
29/4.7
X
1
53/5.8
45/5.6
X
3
20853413
116/6.4
100/6.5
6678559 6755963 6755965
90/5.1 31/8.6 32/7.4
95/5.1 29/8.6 33/7.9
generated list of proteins for those with a relationship to AD. Islet chaperone proteins identified in this studied that have also been examined in AD include GRP78, valosin-containing protein, calreticulin, protein disulfide isomerase, DnaK, HSP70, HSP60, and TCP-1. Islet neural proteins identified in this study and associated with AD include proprotein convertase subtilisin, collapsin response mediator protein 2 (CRMP-2), ubiquinolcytochrome c reductase core protein, L-3-hydroxyacyl-Coenzyme A dehydrogenase, glutamine synthetase, peroxiredoxin (1 and 5), and secretogogin. Aggregates of misfolded proteins, such as AD amyloid, likely arise because some intermediate structures, formed during the folding process, briefly expose hydrophobic regions on their surfaces, and these bind to similar hydrophobic surfaces in nearby folding molecules instead of becoming incorporated into the appropriate structure.11 Endoplasmic reticulum (ER) chaperones have been identified as central components of the quality-control network responsible for correct protein trafficking and for responding to apoptotic and oxidative stress. A currently popular hypothesis is that impaired chaperone function leads to amyloid deposition in tissues.11,12 Some chaperone proteins identified in this study are common to multiple tissue types including liver, muscle, and adipose tissue (i.e., GRP78, calreticulin, protein disulfide isomerase, GRP60, and T-complex
X
1 X
3 4 1
protein 1).7 Valosin-containing protein, also found in this study, has not been described in islets or pancreas previously. Misfolded or partially folded proteins have a strong propensity to aggregate, and members of the HSP60 and HSP70 families may be the most critical chaperones for preventing this activity. The HSP70 chaperone, DnaK, hinders protein aggregation induced by Alzheimer’s β amyloid fibril.13 Protein disulfide isomerase activity speeds the formation of disulfide bonds and inhibits protein aggregation. Misfolded proteins are translocated from the ER, polyubiquinated, and transported to cytosolic proteosomes for degradation. A chaperone identified in the current study, valosin-containing protein, has important transport activity and prevents the accumulation of ubiquinated proteins.14 Improper retranslocation of proteins to proteasomes is thought to be important in the pathogenesis of AD.15 In vitro, mutation of valosin-containing protein results in abnormal aggregation, neural cell death and a pathology resembling that of AD.16 Cellular stresses, such as reactive oxygen species, likely play a major role in the development of AD.17 Chaperones serve an important protective role against these insults. For example, in experimental models of AD that utilize either Aβ-peptide with ferrous iron or peroxide, HSP70 and GRP78 reduce neuronal injury.18,19 Amyloid deposition in AD may involve the inability of chaperone proteins to keep pace Journal of Proteome Research • Vol. 2, No. 2, 2003 203
research articles with the metabolic activity of stressed cells. Prolonged ER stress leads to cell death, decreased chaperones, and is linked to the pathogenesis of AD.12 Chaperones identified in this study, including GRP78, HSP-70, HSP-60, TCP-1, protein disulfide isomerase, and calreticulin, may be differentially expressed or unchanged in neurodegenerative diseases.20-27 In addition to the AD-associated chaperone proteins, neural proteins also implicated in the pathogenesis of AD were found in pancreatic islets. These proteins include proprotein convertase subtilisin, CRMP-2, ubiquinol-cytochrome c reductase core protein, L-3-hydroxyacyl-Coenzyme A dehydrogenase, glutamine synthetase, peroxiredoxin (1 and 5), and secretogogin. It is well established that islets are richly innervated and have intrinsic neural activity.28 Unlike chaperones, most neurally associated proteins identified in this study are less widely distributed among various tissue types (e.g., proprotein convertase, CRMP2, secretagogin). However, some neurally associated proteins found in this study (e.g., ubiquinol - cytochrome c reductase core protein 1) will likely be characterized with a wider distribution among tissues as genomic and proteomic investigations proceed. The neuroendocrine enzyme, proprotein convertase is an endopeptidase that belongs to a family of proteins that have been directly implicated in the most critical pathway of AD pathogenesis: cleavage of β-secretase into its mature form that is responsible for generating the Aβ protein believed to cause AD.29 Secretogogin, a marker of neuronal damage, is another neuroendocrine protein that is highly expressed in islets and in the hippocampus, an area of the brain among the earliest and most extensively damaged regions in AD.30 Neural trophic factors guide the growth of neuronal axons and dendrites. CRMP-2, a neural trophic mediator, has not been previously reported in pancreatic islets or endocrine tissue. CRMP-2 mediates signals for semaphorins, a large family of proteins that regulate axonal guidance in the developing nervous system. CRMP-2 has been identified in neurofibrillary tangles in AD brains.31 Precipitation of CRMP-2 onto neurons may lead to poor neuritic and/or axonal guidance signals and the generation of tangle-bearing neurons eventuating in an acceleration of neuritic degeneration in AD. Defects in the mitochondrial electron transport chain correlate with the oxidative pathology featured in certain neurodegenerative diseases including AD.32 In this context, ubiquinol - cytochrome c reductase core protein 1 and L-3-Hydroxyacylcoenzyme A dehydrogenase have been implicated in the pathogenesis of AD.33,34 Loss of activity of glutamine synthetase in glutamatergic areas of the brain has been implicated in the pathogenesis of AD.35 Another protective protein in neural tissue, peroxiredoxin has antioxidant properties associated with the removal of cellular peroxides. Peroxiredoxin is highly expressed in tissues and cells at risk for diseases related to reactive oxygen species such as AD and altered expression may relate to the pathogenesis of AD.36 None of these neurally associated proteins have been well characterized in islets previously. The relevance of AD-related proteins in islets is unclear. However, it is intriguing to note the similarities and possible relationship between type 2 diabetes, which is a disease that involves islets, and AD. Type 2 diabetes and AD are linked in several clinical studies that show that the incidence of AD is between 2- and 5-fold higher in patients with type 2 diabetes.37-39 Increased dietary fat and a sedentary life style are known risk factors for type 2 diabetes,40 and are probably risk factors for 204
Journal of Proteome Research • Vol. 2, No. 2, 2003
Nicolls et al.
AD.41 Other neurodegenerative diseases such as Huntington’s disease, Friedrich’s ataxia, Werner’s disease, and myotonic dystrophy are all associated with the development of type 2 diabetes.42-45 Both type 2 diabetes and AD are disorders of senescence and are the two most prevalent diseases characterized by localized (as opposed to systemic) amyloid deposition.3,46,47 Amyloid deposits consist of small proteins that form fibrils with a β-pleated structure.3 Islet and AD amyloid may develop due to altered secretion and/or processing of IAPP and Aβ, respectively.48,49 IAPP is a soluble 37-amino acid peptide, cleaved from its precursor form by prohormone convertases, processed in the Golgi, stored in secretory granules and is normally secreted by β-cells into the circulation.3,50 Aβ protein, in contrast to IAPP, is an insoluble 40-42 amino acid peptide that is cleaved from its precursor form by R-, β-, and γ- secretases and is released into the extracellular space.51 Misfolded IAPP and Aβ protein become β-pleated sheets associating with serum amyloid P component, apolipoprotein E and heparan sulfate proteoglycan.2,49 IAPP and Aβ protein share 38% sequence similarity, are directly toxic to surrounding tissue, and activate proinflammatory responses.52-54 IAPP, normally co-secreted with insulin by islet β cells, is also expressed by sensory neurons and has high affinity binding sites in brain.55,56 In several different cell types, elevated levels of either Aβ protein or IAPP cause cell death by altering the homeostasis of free [Ca2+].57 Elevations in intracellular free calcium arising from a loss of homeostasis may be a common mechanism of cellular toxicity in these different forms of amyloid deposition. Of interest, a transgenic mouse model of AD in which human Aβ is overexpressed actually develops widespread pancreatic (rather than brain) acinar amyloid deposits that worsen with age.58 Curiously, the profoundly increased pancreatic levels of Aβ did not correlate with an increase in transgenic mRNA, suggesting that pancreatic cells are in some way specifically sensitive to amyloid deposition.
Conclusions In addition to identifying some of the major protein constituents of the normal mouse islet, this proteomic analysis revealed several islet proteins that have been studied in AD. Loss of chaperone activity may result in the accumulation of protein aggregates and amyloid deposits that are locally toxic to both islet cells (type 2 diabetes) and neural cells (AD). That chaperones are highly expressed in mouse adult pancreatic islets underscores the likely importance of the protein quality control network in these metabolically active cells. In addition, this study identified neural proteins not previously identified in pancreatic islets with clear associations to neurodegenerative diseases including AD. The preponderance of proteins associated with AD in islets warrants speculation about the potential parallel between the pathophysiology of type 2 diabetes and AD. For example, the cellular milieu in islet and brain tissue, where neuropeptidergic activity is high, may favor the localized pattern of amyloid deposition featured in type 2 diabetes and AD. If the pathogenesis of type 2 diabetic pancreatic lesions is likened to that of AD, numerous models of disease can be tested. Future avenues of investigation will focus on the effects of chaperones and neural proteins on the onset of disease in experimental models of type 2 diabetes and on the presence of these same proteins in diseased pancreatic tissue. When rigorous demands are placed on protein identification criteria to minimize false positives, and the findings are thoughtfully
research articles
Alzheimer’s-Related Proteins in Islets
reviewed in a biological context, proteomics offers considerable potential as a hypothesis-generating tool for the biomedical scientist. This study serves as an example of how the consideration of the data generated by proteomics can be utilized to enhance of our understanding of biological systems.
Acknowledgment. This work was support by NIDDK Grant Nos. P30 DK57516 (M.R.N., J.C.H., M.W.D.) and DK33470 (R.G.G.). The authors wish to thank Dr. Kim Fung, Dr. Srdjan Askovic, Ms. Nanette Gomez, Ms. Traci Lyons, and Dr. Seema Kotari for excellent technical assistance. We are also grateful to Drs. Steven Kahn, Marilyne Coulombe, Jane Reusch, Dale Wegmann, and Marty Zamora for their critical insights and review to the development of this manuscript. References (1) Scearce, L. M.; Brestelli, J. E.; McWeeney, S. K.; Lee, C. S.; Mazzarelli, J.; Pinney, D. F.; Pizarro, A.; Stoeckert, C. J., Jr.; Clifton, S. W.; Permutt, M. A.; Brown, J.; Melton, D. A.; Kaestner, K. H. Diabetes 2002, 51, 1997-2004. (2) Hayden, M. R.; Tyagi, S. C. JOP 2001, 2, 124-39. (3) Kahn, S. E.; Andrikopoulos, S.; Verchere, C. B. Diabetes 1999, 48, 241-53. (4) Hyman, B. T.; Trojanowski, J. Q. J. Neuropathol. Exp. Neurol. 1997, 56, 1095-7. (5) Nicolls, M. R.; Coulombe, M.; Yang, H.; Bolwerk, A.; Gill, R. G. J. Immunol. 2000, 164, 3627-34. (6) Li, G.; Waltham, M.; Anderson, N. L.; Unsworth, E.; Treston, A.; Weinstein, J. N. Electrophoresis 1997, 18, 391-402. (7) Sanchez, J. C.; Chiappe, D.; Converset, V.; Hoogland, C.; Binz, P. A.; Paesano, S.; Appel, R. D.; Wang, S.; Sennitt, M.; Nolan, A.; Cawthorne, M. A.; Hochstrasser, D. F. Proteomics 2001, 1, 13663. (8) Sanchez, J. C.; Converset, V.; Nolan, A.; Schmid, G.; Wang, S.; Heller, M.; Sennitt, M. V.; Hochstrasser, D. F.; Cawthorne, M. A. Mol. Cell Proteomics 2002, 1, 509-16. (9) John, N. E.; Andersen, H. U.; Fey, S. J.; Larsen, P. M.; Roepstorff, P.; Larsen, M. R.; Pociot, F.; Karlsen, A. E.; Nerup, J.; Green, I. C.; Mandrup-Poulsen, T. Diabetes 2000, 49, 1819-29. (10) Larsen, P. M.; Fey, S. J.; Larsen, M. R.; Nawrocki, A.; Andersen, H. U.; Kahler, H.; Heilmann, C.; Voss, M. C.; Roepstorff, P.; Pociot, F.; Karlsen, A. E.; Nerup, J. Diabetes 2001, 50, 1056-63. (11) Ellis, R. J.; Pinheiro, T. J. Nature 2002, 416, 483-4. (12) Muchowski, P. Neuron 2002, 35, 9-12. (13) Konno, T. Biochemistry 2001, 40, 2148-54. (14) Dai, R. M.; Li, C. C. Nat. Cell Biol. 2001, 3, 740-4. (15) Checler, F.; da Costa, C. A.; Ancolio, K.; Chevallier, N.; LopezPerez, E.; Marambaud, P. Biochim. Biophys. Acta 2000, 1502, 1338. (16) Hirabayashi, M.; Inoue, K.; Tanaka, K.; Nakadate, K.; Ohsawa, Y.; Kamei, Y.; Popiel, A. H.; Sinohara, A.; Iwamatsu, A.; Kimura, Y.; Uchiyama, Y.; Hori, S.; Kakizuka, A. Cell Death Differ. 2001, 8, 977-84. (17) Pratico, D. Biochem. Pharmacol. 2002, 63, 563-567. (18) Guo, Z. H.; Mattson, M. P. Exp. Neurol. 2000, 166, 173-9. (19) Paschen, W.; Mengesdorf, T.; Althausen, S.; Hotop, S. J. Neurochem. 2001, 76, 1916-24. (20) Martin, J. E.; Swash, M.; Mather, K.; Leigh, P. N. J. Neurol. Sci. 1993, 118, 202-6. (21) Brenneman, D. E.; Gozes, I. J. Clin. Invest. 1996, 97, 2299-307. (22) Yoo, B. C.; Seidl, R.; Cairns, N.; Lubec, G. J. Neural Transm. Suppl. 1999, 57, 315-22. (23) Taguchi, J.; Fujii, A.; Fujino, Y.; Tsujioka, Y.; Takahashi, M.; Tsuboi, Y.; Wada, I.; Yamada, T. Acta Neuropathol. (Berl) 2000, 100, 15360. (24) Kim, H. T.; Russell, R. L.; Raina, A. K.; Harris, P. L.; Siedlak, S. L.; Zhu, X.; Petersen, R. B.; Shimohama, S.; Smith, M. A.; Perry, G. Antioxid. Redox. Signal 2000, 2, 485-9.
(25) Cummings, C. J.; Sun, Y.; Opal, P.; Antalffy, B.; Mestril, R.; Orr, H. T.; Dillmann, W. H.; Zoghbi, H. Y. Hum. Mol. Genet. 2001, 10, 1511-8. (26) Schuller, E.; Gulesserian, T.; Seidl, R.; Cairns, N.; Lube, G. Life Sci. 2001, 69, 263-70. (27) Yoo, B. C.; Kim, S. H.; Cairns, N.; Fountoulakis, M.; Lubec, G. Biochem. Biophys. Res. Commun. 2001, 280, 249-58. (28) Persson-Sjogren, S.; Forsgren, S.; Taljedal, I. B. Peptides 2000, 21, 741-52. (29) Pinnix, I.; Council, J. E.; Roseberry, B.; Onstead, L.; Mallender, W.; Sucic, J.; Sambamurti, K. Faseb J. 2001, 15, 1810-2. (30) Gartner, W.; Lang, W.; Leutmetzer, F.; Domanovits, H.; Waldhausl, W.; Wagner, L. Cereb. Cortex 2001, 11, 1161-9. (31) Yoshida, H.; Watanabe, A.; Ihara, Y. J. Biol. Chem. 1998, 273, 9761-8. (32) Cassarino, D. S.; Bennett, J. P., Jr. Brain Res. Brain Res. Rev. 1999, 29, 1-25. (33) Yan, S. D.; Fu, J.; Soto, C.; Chen, X.; Zhu, H.; Al-Mohanna, F.; Collison, K.; Zhu, A.; Stern, E.; Saido, T.; Tohyama, M.; Ogawa, S.; Roher, A.; Stern, D. Nature 1997, 389, 689-95. (34) Kim, S. H.; Vlkolinsky, R.; Cairns, N.; Lubec, G. Cell Mol. Life Sci. 2000, 57, 1810-6. (35) Arias, C.; Becerra-Garcia, F.; Tapia, R. Neurobiology 1998, 6, 3343. (36) Kim, S. H.; Fountoulakis, M.; Cairns, N.; Lubec, G. J. Neural Transm. Suppl. 2001, 61, 223-35. (37) Ott, A.; Breteler, M. M.; van Harskamp, F.; Stijnen, T.; Hofman, A. Am. J. Epidemiol. 1998, 147, 574-80. (38) Ott, A.; Stolk, R. P.; van Harskamp, F.; Pols, H. A.; Hofman, A.; Breteler, M. M. Neurology 1999, 53, 1937-42. (39) Peila, R.; Rodriguez, B. L.; Launer, L. J. Diabetes 2002, 51, 125662. (40) Diabetes Care 2002, 25, 742-9. (41) Mattson, M. P. Brain Res. 2000, 886, 47-53. (42) Podolsky, S.; Leopold, N. A.; Sax, D. S. Lancet 1972, 1, 1356-8. (43) Moxley, R. T.; Corbett, A. J.; Minaker, K. L.; Rowe, J. W. Ann. Neurol. 1984, 15, 157-62. (44) Robinson, S.; Kessling, A. Baillieres Clin. Endocrinol. Metab. 1992, 6, 867-98. (45) Ristow, M.; Giannakidou, E.; Hebinck, J.; Busch, K.; Vorgerd, M.; Kotzka, J.; Knebel, B.; Mueller-Berghaus, J.; Epplen, C.; Pfeiffer, A.; Kahn, C. R.; Doria, A.; Krone, W.; Mueller-Wieland, D. Diabetes 1998, 47, 851-4. (46) Westermark, P. Ups. J. Med. Sci. 1972, 77, 91-4. (47) Glenner, G. G.; Wong, C. W. Biochem. Biophys. Res. Commun. 1984, 120, 885-90. (48) Badman, M. K.; Shennan, K. I.; Jermany, J. L.; Docherty, K.; Clark, A. FEBS Lett. 1996, 378, 227-31. (49) Pepys, M. B. Philos. Trans R Soc. Lond. B Biol. Sci. 2001, 356, 203-10; discussion 210-1. (50) Jaikaran, E. T.; Clark, A. Biochim. Biophys. Acta 2001, 1537, 179203. (51) Esler, W. P.; Wolfe, M. S. Science 2001, 293, 1449-54. (52) May, P. C.; Boggs, L. N.; Fuson, K. S. J. Neurochem. 1993, 61, 2330-3. (53) Tucker, H. M.; Rydel, R. E.; Wright, S.; Estus, S. J. Neurochem. 1998, 71, 506-16. (54) Yates, S. L.; Burgess, L. H.; Kocsis-Angle, J.; Antal, J. M.; Dority, M. D.; Embury, P. B.; Piotrkowski, A. M.; Brunden, K. R. J. Neurochem. 2000, 74, 1017-25. (55) Beaumont, K.; Kenney, M. A.; Young, A. A.; Rink, T. J. Mol. Pharmacol. 1993, 44, 493-7. (56) Mulder, H.; Leckstrom, A.; Uddman, R.; Ekblad, E.; Westermark, P.; Sundler, F. J. Neurosci. 1995, 15, 7625-32. (57) Kawahara, M.; Kuroda, Y.; Arispe, N.; Rojas, E. J. Biol. Chem. 2000, 275, 14 077-83. (58) Kawarabayashi, T.; Shoji, M.; Sato, M.; Sasaki, A.; Ho, L.; Eckman, C. B.; Prada, C. M.; Younkin, S. G.; Kobayashi, T.; Tada, N.; Matsubara, E.; Iizuka, T.; Harigaya, Y.; Kasai, K.; Hirai, S. Neurobiol. Aging 1996, 17, 215-22.
PR025576X
Journal of Proteome Research • Vol. 2, No. 2, 2003 205
