Abstract: Abnormal glucose tolerance and deviant blood glucose levels are late stage clinical parameters that signify diabetes mellitus. To be able to diagnose the disease at an earlier stage and develop new tools for beta cell imaging, new molecular markers are needed. In the present study, five proteins highly expressed in pancreatic islets with no expression in the surrounding exocrine glandular cells of pancreas, and one protein with the opposite expression pattern, were identified by searches in the Human Protein Atlas (www.proteinatlas.org). The proteins were analyzed immunohistochemically on a specially designed tissue microarray, containing isolated human islets and pancreatic tissues with different characteristics, and compared to the expression of previously known markers of endocrine and exocrine pancreatic cells. Of the five novel endocrine markers, tetraspanin-7 was identified as a membrane-bound protein with exclusive positivity in islet cells. Also β-2-microglobulin and ubiquitin carboxyl-terminal hydrolase isozyme L1 were expressed in a majority of islet cells, whereas sad1/unc-84 domain-containing protein 1 and beta-1,3-glucuronyltransferase 1 were positive in a smaller subset of islet cells. The potential exocrine marker galectin-2 was expressed in both exocrine acinary cells and pancreatic ductal cells, with no or low positivity in islet cells. In conclusion, antibody-based proteomics and specially designed tissue microarrays enable identification and exploration of novel proteins with differential expression in pancreatic islets. Here we describe 5 candidate proteins for further investigation of their physiological role and potential involvement in the pathogenesis of diabetes. One of these proteins, tetraspanin-7, is expressed on the cell membrane and could thus be a potential candidate for future development of tracers for beta cell imaging.
Diabetes mellitus is diagnosed on the basis of blood glucose concentration and abnormal glucose tolerance in individuals, a situation not changed over the last 100 years. However, these clinical parameters are late stage markers for the disease and provide little information as to the changes in pancreatic beta cell function preceding the clinical manifestation of the disease. Identification and exploration of novel proteins with differential expression in pancreatic islets, and analysis of their physiological role and potential involvement in underlying mechanisms of beta cell malfunction/destruction, would be of significant clinical advantage. Understanding the events that occur in pancreatic beta cell mass before the onset of type 1 and type 2 diabetes mellitus could possibly allow early intervention strategies to delay or even prevent the onset of disease.
Recent advances in non-invasive imaging technologies such as Magnetic Resonance Imaging (MRI) and Positron Emission Tomography (PET) are likely to provide an opportunity to monitor pancreatic beta cell mass in humans. A range of candidate proteins and metabolites enabling imaging of the islets of Langerhans have received significant interest over the past ten years (Schneider, 2008). However, many of these candidates have lacked the specificity required for in situ imaging of the pancreatic beta cells. The most important issue is the identification of a beta cell target that is not significantly expressed in other abdominal tissues or the exocrine pancreas. Furthermore, an essential requirement for imaging is that the expression level of the protein must be high enough to obtain a good signal-to-noise ratio.
Antibody-based proteomics provides a strategy for the systematic generation and usage of specific antibodies to explore the proteome (Uhlen and Ponten, 2005). Based on such a strategy, the Swedish Human Protein Atlas (HPA) program has been set up to generate a comprehensive map of protein expression patterns in human tissues and cells (Uhlen et al., 2005). This multi-disciplinary research program combines large-scale generation of validated antibodies (Nilsson et al., 2005) with protein profiling in human tissues and cells, using high-throughput immunohistochemistry (Warford et al., 2004) on tissue microarrays (TMAs) (Kononen et al., 1998). The expression of over 8,400 unique proteins corresponding to 42% of the approximately 20,000 protein encoding genes in the human genome (Clamp et al., 2007) have so far been successfully characterized and published on the HPA portal (www.proteinatlas.org) (Berglund et al., 2008). The structure and contents of this database allow for searches and identification of proteins expressed in specified tissues (Bjorling et al., 2007), and thus provide an attractive starting point for further analysis of identified proteins using other assays and more targeted patient material.
In the present investigation, the HPA portal was used to search for proteins with selective expression pattern in pancreas. Based on the immunohistochemical staining pattern, reliability of the antibody, and previous published data, six proteins were selected for further evaluation in a specially designed TMA containing tissues from isolated human islets exposed to various substances in vitro, and pancreatic tissues from normal and diabetic subjects.
Materials and Methods
Culturing of pancreatic islets
All human studies were approved by local ethics committees. Once legal consent had been obtained, pancreatic tissues were procured from multiorgan donors within the Nordic Network for Clinical Islet Transplantation. Intraductal enzyme perfusion, automated digestion-filtration, islet continuous gradient purification, and subsequent islet culture were performed as previously described in detail (Goto et al., 2004).
The islets were cultured for 3 days in culturing bags together with 100 ml medium (CMRL 1066 AppliChem with 10% ABO compatible human serum and additives) in four different glucose and cytokine conditions. Detailed descriptions of the culturing conditions are listed in Table 1.
In vitro cultured pancreatic islets were harvested, fixed in formalin, and dispersed into agarose cell gels for subsequent histoprocessing and paraffin embedding. The paraffin blocks of islet preparations along with pancreatic tissues from both type 2 diabetes mellitus patients and non-diabetic subjects representing a wide variety of features, were used for production of a TMA. The TMA was generated essentially as previously described (Kampf et al., 2004; Andersson et al., 2006), including three tissue cores from each donor block. Descriptions of the characteristics of the different pancreatic tissues are listed in Table 2.
Antibodies directed towards chromogranin A, insulin, glucagon, and somatostatin were selected for basic characterization of subpopulations of endocrine cells in pancreatic islets, and amylase and cytokeratin-19 (KRT19) antibodies were used to visualize the exocrine pancreatic cell populations. The antibody recognizing the pancreas/duodenum homeobox protein 1 (PDX1) was used as a differentiation/maturation marker (Sander and German, 1997).
Using the advanced search function in the Protein Atlas (Bjorling et al., 2007), 110 protein expression patterns were identified, displaying strong positivity in pancreatic islet cells, but no expression in exocrine glandular cells of pancreas. The search result is summarized in Table 3. From this list, 5 proteins were selected to be included in the present study, based on the immunohistochemical staining pattern, reliability of the antibody, and previous published data. Two proteins [sad1/unc-84 domain-containing protein 1 (SUNC1) and tetraspanin-7 (TSPAN7)] displayed a cell type specific expression pattern with positivity essentially restricted to pancreatic islet cells. In addition to expression in pancreatic islet cells, 3 proteins [β-2-microglobulin precursor (B2M), ubiquitin carboxyl-terminal hydrolase isozyme L1 (UCHL1), and beta-1,3-glucuronyltransferase 1 (B3GAT1)] were more ubiquitously expressed in various tissues, with B3GAT1 being expressed in a subset of the islet cells.
A search for proteins with an opposite expression pattern in pancreas, high expression in exocrine glandular cells, and no expression in islet cells identified 166 potential exocrine markers. One protein, galectin-2 (LGALS2), was selected for inclusion in the present study. Detailed information about the manufacturer and dilution of all used primary antibodies are summarized in Table 4.
The TMA sections were immunohistochemically stained essentially as previously described (Paavilainen et al., 2008). In brief, glass slides were baked in 60°C for 45 min, deparaffinized in xylene, hydrated in graded alcohols, and blocked for endogenous peroxidase in 0.3% hydrogen peroxide. For antigen retrieval, slides were boiled in Target Retrieval Solution (Dako, Glostrup, Denmark) using a Decloaking chamber (Biocare Medical, Walnut Creek, CA, USA), then the automated immunohistochemistry was performed using an AutoStainer XL ST5010 (Leica Microsystems GmbH, Wetzlar, Germany). The slides were incubated for 30 min at room temperature with primary antibodies. For detection, the secondary reagent anti-rabbit/mouse HRP-conjugated UltraVision (Thermo Fischer Scientific, Fremont, CA, USA) was used. Following washing steps, the slides were developed with diaminobenzidine as chromogen. Mayers hematoxylin (Sigma-Aldrich, St. Louis, MO, USA) was used as counterstaining. Slides were mounted with Pertex (Histolab AB, Gothenburg, Sweden).
Annotation of immunohistochemically stained images
The immunohistochemically stained TMA slides were scanned using Aperio ScanScope XT, generating high resolution digital images. The images were used to manually analyze the patterns of immunoreactivity in islets from both pancreatic tissues and from isolated in vitro cultures. The fraction of positive cells for each antibody and tissue spot was scored using 10%-intervals.
To further validate the antibodies, the 7 antibodies targeting proteins previously characterized in pancreas were analyzed with Western blotting using 4 lysates from human pancreatic tissues, including 2 preparations of isolated pancreatic islets and 2 exocrine pancreatic tissues, of which 1 islet lysate and 1 exocrine lysate were treated with protease inhibitor. The 6 less known antibodies were tested using 5 lysates — 2 preparations of isolated pancreatic islets, 2 exocrine pancreatic tissues, and 1 whole pancreatic tissue.
Statistical evaluation was performed using the Statistical Package for the Social Sciences (SPSS, Chicago, IL, USA). Results are expressed as mean ± standard deviation. Unpaired t tests were used to test for between-group differences in fraction of positive cells, with p < 0.05 being considered significant.
An in silico discovery strategy based on the advanced search tool (Bjorling et al., 2007) of the HPA portal was used to find proteins with selective expression in the endocrine cells of human pancreas. A search for proteins that are highly expressed in pancreatic islets but not expressed in the exocrine parenchyma resulted in a list of 110 proteins, including proteins with previously well-known functions as well as unknown proteins (Table 3). Five proteins — B3GAT1, B2M, SUNC1, TSPAN7, and UCHL1 — not previously characterized in pancreatic islet cells, as well as LGALS2, a potential marker for exocrine glandular cells, were selected for more in-depth protein profiling.
Basic characterization of pancreatic islets
A specially designed TMA was generated including pancreatic tissues from healthy individuals and patients with type 2 diabetes mellitus, as well as isolated and cultured pancreatic islets exposed to various concentrations of glucose and cytokines in vitro. Using immunohistochemistry, the expression pattern of chromogranin A, insulin, glucagon, somatostatin, and PDX1 was determined in pancreatic tissues and isolated islet preparations (Figure 1) by manual microscopic evaluation.
For chromogranin A, insulin, and somatostatin, distinct cytoplasmic immunoreactivity with a homogenous pattern was observed both in islets in pancreatic tissues and in in vitro cultured islets. The staining pattern of glucagon was heterogeneous and more diffuse in the in vitro cultured islets compared to pancreatic tissue islets, with a larger fraction of the cells displaying weak to moderate cytoplasmic immunoreactivity, in addition to the subset of cells with strong immunoreactivity. PDX1 was strongly stained in nuclei of pancreatic islets, but to a lesser extent also in nuclei of exocrine ductal cells. Mean values of positive cells ± standard deviation were compared between islets in pancreatic tissues and in vitro cultured islets. A reduced fraction of positive cells in in vitro cultured islets compared to islets in pancreatic tissues was observed for chromogranin A (100 ± 0.0% vs. 80 ± 19.1%; p < 0,001), insulin (85 ± 8.8% vs. 46 ± 11.9%; p < 0.001), and PDX1 (78 ± 22.6% vs. 36 ± 13.4%; p < 0.001), whereas there was no significant difference in the amount of glucagon (11 ± 11.8% vs. 15 ± 7.4%; p = 0.45) and somatostatin (9 ± 5.2% vs. 9 ± 2.5%; p = 1.0) expressing cells.
Of the exocrine markers (Figure 2), amylase was found to be distinctly positive in exocrine glandular cells of pancreatic tissues. In 8 ± 5.0% of the in vitro cultured islets, strong cytoplasmic immunoreactivity with a granular or dot-like pattern was observed. LGALS2 was strongly expressed in cytoplasm and nuclei in a majority of exocrine glandular cells. In some cells a polarity was observed, and few cells were more weakly stained. In the in vitro cultured islets, 11 ± 2.4% of the cells displayed strong expression of LGALS2. KRT19 exhibited distinct cytoplasmic and membranous immunoreactivity in ductal cells of pancreatic tissues and in 15 ± 5.3% of the in vitro cultured islets. All remaining cells were negative.
Expression pattern of selected proteins
All five novel endocrine markers showed expression in islet cells, both in pancreatic tissues and in in vitro cultured islets (Figure 3). B2M, a secreted protein part of major histocompatibility complex class I molecules, was strongly expressed in 88 ± 8.0% of the islet cells in pancreatic tissues, compared to 83 ± 14.6% (p = 0.18) in the in vitro cultured islets, with stronger immunoreactivity being observed close to the plasma membrane. The number of B2M expressing cells resembled that of chromogranin A expressing ones. B3GAT1 is involved in glycoprotein biosynthesis and displayed distinct cytoplasmic immunoreactivity with a granular or dot-like pattern in only a small fraction of the islet cells, most similar to the pattern of somatostatin expressing cells. In islets of pancreatic tissues, 12 ± 7.4% of the cells were positive, while the fraction of positive cells was significantly smaller (5 ± 2.5%; p = 0.02) in in vitro cultured islets. Another protein with highly selective expression in Langerhans islets was SUNC1, a protein with unknown function. SUNC1 displayed a partly granular cytoplasmic expression pattern in a majority of the pancreatic islet cells. The distribution resembled that of insulin in islets of the pancreatic tissues, with positivity in 69 ± 17.0% of the cells. The amount of positive cells was significantly lower in the in vitro cultured islets (14 ± 4.9%; p < 0.001). TSPAN7 is a transmembrane protein suggested to be involved in cell differentiation (Boismenu et al., 1996) and cell motility (Penas et al., 2000), displaying distinct cytoplasmic immunoreactivity in islet cells with a distribution similar to insulin. The intensity of the TSPAN7 immunostaining was variable in islet cells, ranging from weak to strong, with 85 ± 17.0% of the islet cells in pancreatic tissues being positive. A reduced fraction of positive cells was observed in in vitro cultured islets (51 ± 19.9%; p < 0.001). UCHL1, a cytoplasmic protein involved in regulation of protein degradation, was strongly expressed in a majority of cells in pancreatic islets (80 ± 15.0%) and to a lesser extent in in vitro cultured islets (44 ± 9.2%; p < 0.001). The cytoplasmic staining pattern was homogenous and often accompanied with nuclear immunoreactivity.
For most of the proteins investigated, there was no difference in the level of expressed protein in islets subjected to different culture conditions or in islets of pancreatic tissues from the different groups of patients. Although not significant, some trends were noted despite the few samples representing each group. Islet preparations cultured in high concentration of glucose (16.7 mM) had lower expression of PDX1 and B2M, with 57 ± 23.6% (PDX1) and 63 ± 14.1% (B2M) positive islet cells in high glucose compared to 78 ± 3.5% (PDX1) and 82 ± 2.4% (B2M) in normal glucose. In in vitro cultured islets grown in normal glucose concentration (5.5mM) with supplemented additional cytokines (50U TNF + 50U Interferon γ + 50U IL1β/100ml culturing medium), a higher expression of B2M and UCHL was found. 98 ± 2.4% (B2M) and 47 ± 9.4% (UCHL) of islet cells were positive as compared to 82 ± 2.4% (B2M) and 38 ± 2.4% (UCHL) positive islet cells in islets cultured without the supplement of cytokines. In islets from pancreatic tissues, the expression of B2M and TSPAN7 was reduced in tissues subjected to long time of ischemia (>22 h) as compared to short time of ischemia (<4 h). Prolonged ischemia resulted in 80 ± 9.4% (B2M) and 63 ± 24.8% (TSPAN7) positive islet cells as compared to 97 ± 0.0% (B2M) and 98 ± 2.4% (TSPAN7) in tissues with only short duration of ischemia. There was no obvious tendency towards a difference in protein expression levels between type 2 diabetes mellitus patients and the other donators for any of the analyzed proteins.
To further validate the antibodies used in the study and to examine if the protein expression patterns found in immunohistochemistry also were reflected in Western blot, all antibodies were analyzed with Western blotting, using lysates from isolated endocrine- and exocrine pancreatic cells (Figure 4). Of the 7 proteins previously well-known to be expressed in pancreas, bands of predicted size were found for insulin, somatostatin, amylase, and KRT19. Chromogranin A displayed bands of predicted size, but also several additional bands of non-predicted sizes. Glucagon and PDX1 displayed larger bands than the predicted sizes, which for PDX1 was expected according to previous results (Li et al., 2010). The addition of protease inhibitor to the tissues did not result in any difference for any of the antibodies (Figure 4a).
Of the antibodies corresponding to the 6 newly identified proteins with differential expression in pancreas, B2M, UCHL1, and LGALS2 detected proteins of predicted size, with UCHL1 exhibiting strong bands in lysates from islet cells and LGALS2 displaying strong bands in lysates from exocrine cells. B2M showed bands of predicted size both in wells representing islet cells and in wells representing exocrine cells. SUNC1 displayed distinct bands smaller than predicted in lysates from islet cells and in one of the two lysates from exocrine cells. TSPAN7 exhibited strong bands in close vicinity to the loading well, only in lysates from islet cells. No bands were seen for B3GAT1.
The discrepancies between immunohistochemistry and Western blotting for certain antibody-protein interactions are not unexpected, as protein epitopes are altered in different ways dependent on effects of denaturing, e.g., formalin and SDS-PAGE. The reason for TSPAN7 to be remaining near the loading wells is unclear and may be due to complex binding or modifications of the TSPAN7 protein. Dilution of the lysate did not alter the ability of TSPAN7 to move further into the gel.
The identification of specific gene expression patterns is a major challenge to increase our understanding of normal islet function and the pathogenesis of diabetes. The discovery of new proteins with a selective expression in Langerhans islets as compared to surrounding exocrine pancreas also provides a starting point for the identification and development of candidates to determine beta cell mass. Tissue-restricted transcripts have been characterized by various groups, and efforts using cDNA libraries have established a relative abundance of >2,000 islet transcripts, including both well-known and potential new markers (Cras-Meneur et al., 2004). Alternative efforts using oligonucleotide chips have also been employed to uncover markers of healthy or diseased islet cell masses (Maffei et al., 2004). Furthermore, genes of pancreatic islets modified by viral infections and cytokines have been described (Ylipaasto et al., 2005). In order to understand the correlation between genotype and phenotype, an important complement to transcript profiling is to determine protein profiles in a tissue context. Two-dimensional gel electrophoresis and mass spectrometry have been used to generate reference maps of the human pancreatic islet proteins or peptides as a resource for future analyses of pancreatic islet biology (Ahmed et al., 2005; Metz et al., 2006). Further development of mass spectrometry-based technologies to overcome sensitivity limitations will most probably become increasingly important in the search for proteins involved in beta cell biology.
An alternative proteomics approach is antibody-based proteomics (Uhlen and Ponten, 2005), which relies on immunohistochemistry and prevails as an invaluable method for in situ visualization of protein expression patterns. Although immunohistochemistry is not a quantitative method, it allows for the detection and localization of defined proteins in a tissue context at cellular or subcellular resolution, provided that protein-specific antibodies are available. The HPA program has employed a strategy to generate antibodies towards human proteins on a global scale and to use these to create a comprehensive atlas of protein expression patterns in human normal and cancer tissues as well as in cell lines. In addition to protein profiling using immunohistochemistry and brightfield microscopy, confocal microscopy with fluorescently labeled antibodies is also used to provide a more detailed analysis and higher resolution of the subcellular localization pattern of each protein (Barbe et al., 2008). This resource with over 9 million annotated images can be used to identify proteins with cell- and tissue type specific expression patterns, and to detect various types of biomarkers using in silico based methods (Bjorling et al., 2007). A large fraction of the human genome is expressed on the protein level in any given cell type and protein expression restricted to only a single cell type is very uncommon (Ponten et al., 2009). However, when searching for proteins expressed in a certain cell type as compared to other surrounding cell types within a defined organ, e.g., Langerhans islet cells and exocrine epithelial cells in the pancreas, there are substantially more proteins that meet the requirement of being differentially expressed.
The general outline of immunohistochemistry on TMAs is well suited for more targeted studies, and with adequate materials available, any tissue or disease can be studied in extended and more in-depth analyses (Ponten et al., 2008). The use of TMAs, containing tissues representing large patient cohorts coupled to clinical databases, also allows for controlled studies on single slides sparing valuable tissues and using only small amounts of reagent (Kononen et al., 1998). Since all tissues are processed and analyzed simultaneously, experimental variance is minimized, rendering interpretation of results more accurate and robust. In the present study we have used an experimental set-up combining queries in the HPA database with immunohistochemistry and TMA technology, applied on pancreatic islet cells to address questions regarding diabetes and islet cell biology. Based on the immunohistochemical staining pattern, antibody reliability, and previous published data, six proteins distinctly expressed in defined cells of pancreatic tissue were selected for this study. The aim was to determine the expression pattern of the selected proteins in in vitro cultured islet cell preparations exposed to different conditions, and on pancreatic tissues from subjects representing different states of the metabolic syndrome.
The protein expression pattern of B2M and UCHL1, with cytoplasmic positivity in a majority of the islet cells, is consistent with expression corresponding to the beta cells. To establish exclusive expression of these proteins in beta cells, further studies are needed using double labeling techniques and other in vitro assays. Previous studies have revealed that major histocompatibility complex class I molecules, which B2M is part of, are hyperexpressed by the endocrine cells during the pathogenic process of type 1 diabetes mellitus (Foulis, 1996), and B2M is produced at high levels in islets from post-mortem pancreatic tissue in an individual who repeatedly tested positive for islet cell antibodies (Oikarinen et al., 2008). In addition, it is suggested that an intact B2M-pathway is necessary for islet allograft survival in mice (Beilke et al., 2004). Although not significant, the expected physiological processes were reflected when comparing the protein expression in pancreatic islets with different characteristics, e.g., addition of cytokines resulted in altered expression levels of B2M. TSPAN7 was expressed to a similar extent as B2M and UCHL1, although this protein appeared with a hinted polarity in positive islet cells. Provided that the expression of TSPAN7 is exclusively expressed in pancreatic islets and that the protein, which contains transmembrane spanning regions, also is expressed on the cell surface, TSPAN7 may be a potential candidate for future development of tracers for beta cell imaging. LGALS2 displayed a distinct immunohistochemical staining pattern restricted to exocrine cells of pancreas as compared to islets. The LGALS2 staining pattern appeared less diffuse than what was found using an antibody towards amylase, suggesting that LGALS2 could be used as a marker for exocrine glandular cells, independent of the level of exocrine granules within the cells. In general, antibodies recognizing proteins with selective positivity in islet cells displayed a reduced amount of positive islet cells in the in vitro cultured islets compared to the islets in pancreatic tissues. This result is consistent with previous findings of slight impurity of the islet isolations, due to a remaining contribution of exocrine and ductal cells in the cell culture (Shapiro et al., 2000). This was also confirmed with basic characterization of the islet cells with antibodies against amylase, LGALS2, and KRT19.
In conclusion, our study describes an efficient strategy to identify novel protein targets that are specifically expressed in pancreatic islet cells, and gives an example of how the identified targets can be further exploited in a selected set of well-defined tissues and in vitro cultured islets subjected to experimental perturbations. Larger cohorts, deeper characterization, and functional studies are needed to transform the discovery and identification of proteins with islet cell specific expression patterns and to further understand islet cell biology and the development of new candidates for clinical beta cell imaging.
The authors wish to acknowledge all members of the Human Protein Atlas project and the Department of Clinical Immunology, Uppsala University for making this work possible. The project is financially supported by The Wallenberg Research Foundation (KAW), the Swedish Research Council (VR), the Juvenile Diabetes Research Foundation, and Vinnova Proj# 30552-1.
Ahmed M, Forsberg J, Bergsten P. Protein profiling of human pancreatic islets by two-dimensional gel electrophoresis and mass spectrometry. J Proteome Res 4(3):931-40, 2005.
Andersson AC, Stromberg S, Backvall H, Kampf C, Uhlen M, Wester K, Ponten F. Analysis of protein expression in cell microarrays: a tool for antibody-based proteomics. J Histochem Cytochem 54(12):1413-23, 2006.
Barbe L, Lundberg E, Oksvold P, Stenius A, Lewin E, Bjorling E, Asplund A, Ponten F, Brismar H, Uhlen M, Andersson-Svahn H. Toward a confocal subcellular atlas of the human proteome. Mol Cell Proteomics 7(3):499-508, 2008.
Beilke J, Johnson Z, Kuhl N, Gill RG. A major role for host MHC class I antigen presentation for promoting islet allograft survival. Transplant Proc 36(4):1173-4, 2004.
Berglund L, Bjorling E, Oksvold P, Fagerberg L, Asplund A, Szigyarto CA, Persson A, Ottosson J, Wernerus H, Nilsson P, Lundberg E, Sivertsson A, Navani S, Wester K, Kampf C, Hober S, Ponten F, Uhlen M. A genecentric Human Protein Atlas for expression profiles based on antibodies. Mol Cell Proteomics 7(10):2019-27, 2008.
Bjorling E, Lindskog C, Oksvold P, Linne J, Kampf C, Hober S, Uhlen M, Ponten F. A web-based tool for in silico biomarker discovery based on tissue-specific protein profiles in normal and cancer tissues. Mol Cell Proteomics 7(5):825-44, 2008.
Boismenu R, Rhein M, Fischer WH,and Havran WL. A role for CD81 in early T cell development. Science 271(5246):198-200, 1996.
Clamp M, Fry B, Kamal M, Xie X, Cuff J, Lin MF, Kellis M, Lindblad-Toh K, Lander ES. Distinguishing protein-coding and noncoding genes in the human genome. Proc Natl Acad Sci U S A 104(49):19428-33, 2007.
Cras-Meneur C, Inoue H, Zhou Y, Ohsugi M, Bernal-Mizrachi E, Pape D, Clifton SW, Permutt MA. An expression profile of human pancreatic islet mRNAs by Serial Analysis of Gene Expression (SAGE). Diabetologia 47(2):284-99, 2004.
Foulis AK. The pathology of the endocrine pancreas in type 1 (insulin-dependent) diabetes mellitus. Apmis 104(3):161-7, 1996.
Goto M, Eich TM, Felldin M, Foss A, Kallen R, Salmela K, Tibell A, Tufveson G, Fujimori K, Engkvist M, Korsgren O. Refinement of the automated method for human islet isolation and presentation of a closed system for in vitro islet culture. Transplantation 78(9):1367-75, 2004.
Kampf C, Andersson AC, Wester K, Bjorling E, Uhlen M, Ponten F. Antibody-based tissue profiling as a tool for clinical proteomics. Clin Proteomics 1:285-99, 2004.
Kononen J, Bubendorf L, Kallioniemi A, Barlund M, Schraml P, Leighton S, Torhorst J, Mihatsch MJ, Sauter G, Kallioniemi OP. Tissue microarrays for high-throughput molecular profiling of tumor specimens. Nat Med 4(7):844-7, 1998.
Li SW, Sun Y, Donelan W, Tang D, Yu H, Scian J, Yang LJ. Expression, purification, and characterization of recombinant human pancreatic duodenal homeobox-1 protein in Pichia pastoris. Protein Expr Purif 72(2):157-61, 2010.
Maffei A, Liu Z, Witkowski P, Moschella F, Del Pozzo G, Liu E, Herold K, Winchester RJ, Hardy MA, Harris PE. Identification of tissue-restricted transcripts in human islets. Endocrinology 145(10):4513-21, 2004.
Metz TO, Jacobs JM, Gritsenko MA, Fontes G, Qian WJ, Camp DG, 2nd, Poitout V, Smith RD. Characterization of the human pancreatic islet proteome by two-dimensional LC/MS/MS. J Proteome Res 5(12):3345-54, 2006.
Nilsson P, Paavilainen L, Larsson K, Odling J, Sundberg M, Andersson AC, Kampf C, Persson A, Al-Khalili Szigyarto C, Ottosson J, Bjorling E, Hober S, Wernerus H, Wester K, Ponten F, Uhlen M. Towards a human proteome atlas: high-throughput generation of mono-specific antibodies for tissue profiling. Proteomics 5(17):4327-37, 2005.
Oikarinen M, Tauriainen S, Honkanen T, Vuori K, Karhunen P, Vasama-Nolvi C, Oikarinen S, Verbeke C, Blair GE, Rantala I, Ilonen J, Simell O, Knip M, Hyoty H. Analysis of pancreas tissue in a child positive for islet cell antibodies. Diabetologia 51(10):1796-802, 2008.
Paavilainen L, Wernerus H, Nilsson P, Uhlen M, Hober S, Wester K, Ponten F. Evaluation of monospecific antibodies: a comparison study with commercial analogs using immunohistochemistry on tissue microarrays. Appl Immunohistochem Mol Morphol 16(5):493-502, 2008.
Penas PF, Garcia-Diez A, Sanchez-Madrid F, Yanez-Mo M. Tetraspanins are localized at motility-related structures and involved in normal human keratinocyte wound healing migration. J Invest Dermatol 114(6):1126-35, 2000.
Ponten F, Gry M, Fagerberg L, Lundberg E, Asplund A, Berglund L, Oksvold P, Bjorling E, Hober S, Kampf C, Navani S, Nilsson P, Ottosson J, Persson A, Wernerus H, Wester K, Uhlen M. A global view of protein expression in human cells, tissues, and organs. Mol Syst Biol 5:337, 2009.
Ponten F, Jirstrom K, Uhlen M. The Human Protein Atlas — a tool for pathology. J Pathol 216(4):387-93, 2008.
Sander M, German MS. The beta cell transcription factors and development of the pancreas. J Mol Med 75(5):327-40, 1997.
Schneider S. Efforts to develop methods for in vivo evaluation of the native beta-cell mass. Diabetes Obes Metab 10(Suppl 4):109-18, 2008.
Shapiro AM, Lakey JR, Ryan EA, Korbutt GS, Toth E, Warnock GL, Kneteman NM, Rajotte RV. Islet transplantation in seven patients with type 1 diabetes mellitus using a glucocorticoid-free immunosuppressive regimen. N Engl J Med 343(4):230-8, 2000.
Uhlen M, Bjorling E, Agaton C, Szigyarto CA, Amini B, Andersen E, Andersson AC, Angelidou P, Asplund A, Asplund C, Berglund L, Bergstrom K, Brumer H, Cerjan D, Ekstrom M, Elobeid A, Eriksson C, Fagerberg L, Falk R, Fall J, et al.. A human protein atlas for normal and cancer tissues based on antibody proteomics. Mol Cell Proteomics 4(12): 1920-32, 2005.
Uhlen M, Ponten F. Antibody-based proteomics for human tissue profiling. Mol Cell Proteomics 4(4):384-93, 2005.
Warford A, Howat W, Mccafferty J. Expression profiling by high-throughput immunohistochemistry. J Immunol Methods 290(1-2):81-92, 2004.
Ylipaasto P, Kutlu B, Rasilainen S, Rasschaert J, Salmela K, Teerijoki H, Korsgren O, Lahesmaa R, Hovi T, Eizirik DL, Otonkoski T, Roivainen M. Global profiling of coxsackievirus- and cytokine-induced gene expression in human pancreatic islets. Diabetologia 48(8):1510-22, 2005.
[Discovery Medicine, 9(49):565-578, June 2010.]