Different Signaling Pathways Expressed by Chicken Naı¨ve CD4+ T Cells, CD4+ Lymphocytes Activated with Staphylococcal Enterotoxin B, and Those Malignantly Transformed by Marek’s Disease Virus Joram J. Buza*,† and Shane C. Burgess‡,§,|,⊥ Vaccine and Infectious Disease Organization, University of Saskatchewan, Saskatoon, Canada, College of Veterinary Medicine, Institute for Digital Biology, Mississippi Agricultural and Forestry Experiment Station, and Life Sciences and Biotechnology Institute, Mississippi State University, Mississippi, 39762 Received December 12, 2007
Proteomics methods, based on liquid chromatography and tandem mass spectrometry, produce large “shotgun” proteomes that are most appropriately compared not at the level of differentially expressed proteins only but at the more comprehensive level of biological networks and pathways. This is now possible with the emergence of functional annotation databases and tools, databases of canonical pathways and molecular interactions and computational text mining tools. Here, we used shotgun proteomics, and the differential proteomics modeling functionalities available in the Pathwaystudio network modeling program to define the cell physiology of Hodgkin’s disease antigen-overexpressing (CD30hi) CD4+ T cell lymphomas using the unique Marek’s disease (MD) natural animal model. CD30hi lymphoma cells have characteristics of activated T cells but are also fundamentally different from their nontransformed healthy counterparts. We compared the cell physiology of naïve, superantigen-activated and MD-transformed CD4+ T cell proteomes. While the superantigen-activated cells had signaling pathways associated with cell activation, inflammation, proliferation and cell death, the MD-transformed cells had growth factor, cytokine, adhesion, and transcription factor signaling responses associated with oncogenicity, cell proliferation, angiogenesis, motility, and metastasis. Keywords: differential proteomics • shotgun proteomics • CD4+ T cells • Staphylococcus • enterotoxin B • Marek’s disease • lymphoma • Pathwaystudio • MudPIT
Introduction Differential proteomics is important for identifying molecular processes involved in different physiologies or pathologies. In classical gel-based proteomics, proteins that are shared between the samples are almost universally ignored and only differentially expressed proteins are selected for identification by mass spectrometry. The inherent limitations of gel-based proteomics (including low detection sensitivity and linearity, poor solubility of membrane proteins, limited loading capacity of gradient pH strips, poor reproducibility of gels, lowthroughput and low linear range of visualization) have restricted its general applicability.1 The demand for more comprehensive proteome coverage fueled the development of “shotgun proteomics” based on liquid chromatography directly combined with tandem mass spectrometry (2-D LC MS2). Shotgun proteomics in turn created new challenges for proteome comparison,2,3 and like 2-D gels, the technique suffers * To whom correspondence should be addressed. Phone: +1-306-9661515. Fax: +1 306-966-7478. E-mail:
[email protected]. † University of Saskatchewan. ‡ College of Veterinary Medicine, Mississippi State University. § Institute for Digital Biology, Mississippi State University. | Mississippi Agricultural and Forestry Experiment Station, Mississippi State University. ⊥ Life Sciences and Biotechnology Institute, Mississippi State University.
2380 Journal of Proteome Research 2008, 7, 2380–2387 Published on Web 04/16/2008
from low reproducibility.2 However, analyzing the data at the level of biological systems (protein networks, signaling pathways, GO groups, biological functions) can alleviate the problems of “missing” data points between compared conditions.4–6 All proteins from each compared condition, whether or not they are present in both samples, are used to generate signaling pathways or GO functions prior to comparison. This approach is fundamentally valid because most proteins do not function in isolation but are involved in one or more cellular pathways and it is the overall effect on these pathways that results in changes in signaling and biological outcome.7 Furthermore, comparing conditions at this higher “systems biology” level is both more sensitive and more “noise” tolerant.8 Such systems level approaches are complicated by the enormous amounts of protein functional information that must be derived from the literature. However, many computational tools and databases have been developed to curate relationships from published literature and map these into canonical networks and pathways. These tools include the two most widely used commercially available Ingenuity Pathways Analysis (IPA, Ingenuity Systems, Inc.) and PathwayStudio (Ariadne Genomics). Recently, Pathwaystudio and IPA were compared, using a Hutchinson-Gilford progeria syndrome.9 Pathwaystudio and IPA were comparable in terms of functionality and produced similar results. However, while IPA knowledge da10.1021/pr700844z CCC: $40.75
2008 American Chemical Society
research articles
Different Signaling Pathways Table 1. Summary of Signaling Pathways That Are Differentially Expressed in Staphylococcal Enterotoxin B-Activated Chicken Peripheral Blood CD4+ T Cells As Compared to Naïve Chicken Peripheral Blood CD4+ T Cells signaling pathway/pathway group
Materials and Methods
%
Tumor necrosis Factor receptor (TNFR) signaling 27.45 Integrin signaling 21.57 Notch signaling 13.73 Interferon R/β signaling 7.84 Death receptor signaling 3.92 Hepatocyte growth factor signaling 3.92 Interleukin 2 signaling 3.92 Interleukin 6 receptor f CCAAT/enhancer binding protein 1.96 Low density lipoprotein receptor f sterol regulatory element 1.96 binding transcription factor 1 protein tyrosine phosphatase, receptor type, J f catenin 1.96 (cadherin-associated protein), delta 1 inducible T-cell costimulator ligand (B7H2) f CD28 signaling 1.96 Delta-like 1 (DLL1)-NOTCH1 signaling 1.96 Transforming growth factor, beta 1(TGFB1) f CD44 signaling 1.96 Caspases signaling 1.96 G protein I,0 signaling 1.96 CD40L signaling 1.96
tabase is manually curated by biocurators, the Pathwaystudio database is curated electronically using automated text-mining engines that are also available to investigators. The IPA is limited to human, mouse, and rat gene products, but the Pathwaystudio database, because the text mining software is available, is able to annotate proteins from any species. Furthermore, Pathwaystudio is much less expensive and more visually functional. We have used both Ingenuity Pathways Analysis and Pathwaystudio and have also found that the two tools produce comparable pathway analysis results for proteins that can be mapped easily to human, mouse, or rat orthologs.10 However, we found that only Pathwaystudio could be used to easily compare and contrast the pathways or biological functions generated from two or more samples and without the limitation of orthology assignment. Here, we used shotgun proteomics, and the differential proteomics modeling functionalities available in the Pathwaystudio network modeling program, to define the cell physiology of Hodgkin’s disease antigen-overexpressing (CD30hi) CD4+ T cell lymphomas using the Marek’s disease (MD) model. Marek’s disease is caused by the highly infectious MD R-herpesvirus (MDV) and is a unique natural animal model for lymphomas that overexpress CD30, the tumor necrosis factor receptor superfamily member 8.11,12 CD30hi lymphoma cells have characteristics of activated T cells but are also fundamentally different from their nontransformed healthy counterparts. We compared the cell physiology of naïve, superantigen (SEB)activated and the UA-0113 MD-transformed CD4+ T cell proteomes. SEB is an exotoxin produced by Staphylococcus aureus.14 While the SEB-activated cells had signaling pathways associated with cell activation, inflammation, proliferation, and cell death, the neoplastically transformed cells had growth factor, cytokine, adhesion, and transcription factor signaling responses associated with oncogenicity, cell proliferation, angiogenesis, motility, metastasis, and T-regulatory phenotype. As an immune activator, SEB is involved in food poisoning and toxic shock15 and is a category B priority agent that can be used as an air-borne, food-borne, and water-borne weapon.16 Our global proteomic analysis of pathways activated by SEB has also provided a model of its mechanisms of action on T-helper cells.
Naïve, Staphylococcal Enterotoxin B-Activated and MD-Transformed CD4+ T Cells. Blood was drawn from jugular vein of adult chicken using heparinized vacuum tubes (Vacutainer, Beckton Dickinson, NJ). Peripheral blood mononuclear cells (PBMC) were isolated by density gradient centrifugation (20 °C; 400g; 30 min) using Histopaque 1077 (Sigma, St. Louis, MO). The PBMC were divided into two equal aliquots of 10 × 109 cells for isolation of naïve CD4+ T cells and SEB-activated CD4+ T cells, respectively. Naı¨ve CD4+ T lymphocytes were isolated directly from PBMC by magnetic cell sorting using the anti-FITC MACS beads (Miltenyi Biotech, Auburn, CA) and the mouse anti-chicken CD4-FITC (Southern Biotech, Birmingham, AL) according to manufacturer’s instructions. For isolation of SEB activated CD4+ T cells, PBMC were first stimulated with 1 µg/mL SEB (Calbiochem/EMD Biosciences, La Jolla, CA) for 24 h at 41 °C, and subsequently, the CD4+ were isolated using magnetic cell sorting using the anti-FITC MACS beads and the mouse anti-chicken CD4-FITC. The Marek’s disease transformed UA-01 cell line was obtained from Dr. M. Parcells of University of Delaware, and grown as described.13 Flow Cytometry. The activation state of the CD4+ T cells was monitored by flow-cytometry analysis of cell size and MHC class II expression as previously described.11 Single cell suspensions (106 cells) of naïve and SEB-activated PBMC were stained for CD4 and MHC-II using mouse IgG1 anti-chicken CD4 R-phycoerythrin (R-PE) conjugate (Southern Biotech, Birmingam, AL) and mouse IgG1 anti-chicken MHC-II fluorescein (FITC) conjugate (Southern Biotech). Isotype-matched controls used were purchased from ID Laboratories (Ontario, Canada). Cell size and fluorescence was analyzed using the FACS scan flow cytometer (Beckton Dickinsons, San Jose, CA). We calculated the proportions of MHC class II-expressing CD4+ T lymphocytes as well as the proportions of blasts as defined by forward scatter (FSC) and side scatter (SSC) as described previously.11 Differences between triplicate samples of the two cell populations were analyzed by student t test. Protein Extraction. Proteins were extracted in triplicate from the 3 samples (naïve, SEB-activated and MD-transformed CD4+ T cell lymphoma cell line) using differential detergent fractionation (DDF), each producing 4 DDF fractions, and the protein lysates were processed for protein identification exactly as described.10 Protein lysates were precipitated using 25% trifluoroacetic acid (TFA) and resuspended in 0.1 M ammonium bicarbonate and 5% acetonitrile (ACN), and the pH was adjusted to g7.5 using 1 M Tris, pH 8.0. The protein solutions were then reduced using dithiothreitol (final concentration 5 mM; 65 °C, 5 min) and alkylated with iodoacetamide (final concentration 10 mM; 30 °C, 30 min). The proteins were then digested overnight using molecular biology grade porcine trypsin (Promega, Madison, WI; 50:1 final substrate/trypsin ratio; 37 °C). The resulting peptides were desalted using a C18 microtrap (Microm Bioresources, Inc., Auburn, CA) and eluted using 0.1% TFA and 95% ACN solution, vacuum-dried, and resuspended in 0.1% formic acid. Mass spectrometry analysis was done by two-dimensional liquid chromatography electrospray ionization tandem mass spectrometry (2D LC ESI MS2) using a Thermo Separations P4000 quaternary gradient pump LCQ Deca XP Plus (ThermoElectron Corporation; San Jose, CA) as described previously.17 LC was done by strong cation exchange (SCX) followed by reverse phase (RP) LC coupled directly in line with ESI ion trap mass spectrometer. Samples were loaded into a LC gradient ion exchange system (Thermo Journal of Proteome Research • Vol. 7, No. 6, 2008 2381
research articles
Buza and Burgess
Table 2. Summary of Signaling Pathways That Are Differentially Expressed in Marek’s Disease Virus-Transformed Chicken CD4+ T Cell Lymphoma Cell Line As Compared to Naïve Chicken Peripheral Blood CD4+ T Cells category
Soluble factors/soluble factor receptor mediated
subcategory
signaling pathway/pathway group
Growth factors
Tranforming growth factor B/TGFB receptor signaling Fibroblast growth factor/FGFR signaling Met proto-oncogene (hepatocyte growth factor receptor) (MET) Epidermal growth factor/EGFR Nerve growth factor receptor (NGFR) signaling Vascular endothelial growth factor (VEFG) Hepatocyte growth factor HGF Insulin like growth factor (IGF) signaling Leukemia inhibitory factor precursor (LIF) signaling Brain-derived neurotrophic factor precursor (BDNF) signaling Colony-stimulating factor 1 macrophage (CSF-1) Neurotrophin 3/NTF3 signaling Stromal cell-derived factor 1 precursor (SDF-1) Neuregulin 1 (NRGI) Platelet-derived growth factor (DGF) signaling TNFR signaling Interferons G/IFN B signaling Interleukin 6 Inhibin beta A (INHBA) chain Interleukin-4 signaling Oncostatin C-C chemokine receptor type 2 (MCP-1 R) Interleukin 1 Interleukin 2 Interleukin 11 Intreleukin 12 Insulin (INS) signaling Prostaglandin Factor (PGF) signaling Gonadotropin-releasing hormone receptor (GN RHR) signaling Erythropoietin signaling Anti-Muellerian hormone type-2 receptor precursor (AMHR)
7.27
Notch signaling Fos Integrin signaling Proteinase-activated receptor 1 precursor/F2R signaling Activin Receptor (ACVR) signaling Cholecystokinin type receptor (CCKR) signaling Glutamate receptor, metabotropic 2(GRM2 and GRM5) Cannabinoid receptor 1 (CNR1) Dopamine receptor D2 (DRD2) Endothelin receptor type B (EDNRB) Secreted phosphoprotein 1 Others (23 pathways each 0.49%)
5.91 5.00 3.18 3.18
Cytokines
Hormones
Nonsoluble factor mediated
Separations P4000 quaternary gradient pump coupled with a 0.32 × 100 mm BioBasic strong cation exchange column). A flow rate of 3 µL/min was used for both SCX and RP columns. A salt gradient was applied in steps of 0, 5, 10, 15, 20, 25, 30, 35, 40, 45, 50, 57, 64, 71, 79, 90, 110, 300, and 700 mM ammonium acetate in 5% ACN and 0.1% formic acid, and the 2382
Journal of Proteome Research • Vol. 7, No. 6, 2008
%
% subcategory
% category
37.73
62.73
5.45 4.09 3.64 3.64 3.18 1.82 1.82 1.36 0.91 0.91 0.91 0.91 0.91 0.91 5.45 3.64 2.27 1.82 1.36 1.36 1.36 0.91 0.91 0.45 0.45 1.82 0.91 0.91
20.00
5.00
0.45 0.91
37.27
2.27 1.82 0.91 0.91 0.91 0.91 0.91 11.36
resultant peptides were loaded directly into the sample loop of a 0.18 × 100 mm BioBasic C18 RPLC column (ThermoElectron). The reverse phase gradient used 0.1% formic acid in ACN and increased the ACN concentration in a linear gradient from 5% to 30% in 30 min and then 30% to 65% in 9 min followed by 95% for 5 min and 5% for 15 min. The spectrum collection
research articles
Different Signaling Pathways Table 3. Summary of Signaling Pathways That Are Unique for Staphylococcal Enterotoxin B-Activated Chicken Peripheral Blood CD4+ T Cells As Compared to Naïve Chicken Peripheral Blood CD4+ T Cells and Marek’s Disease Virus-Transformed Chicken CD4+ Lymphoma T Cell Line pathway/pathway group
%
Tumor necrosis factor receptor (TNFR) signaling Integrin signaling Interleukin 2 /Interleukin 2 receptor Interleukin 6 Death receptor Caspases Notch homologue 1, translocation-associated (NOTCH1) Low density lipoprotein signaling pathway Protein tyrosine phosphatase signaling pathway G protein signaling CD40L signaling
43.8 25.0 6.3 3.1 3.1 3.1 3.1 3.1 3.1 3.1 3.1
time was 59 min for every strong cation exchange step. The mass spectrometer was configured to optimize the duty cycle length with the quality of data acquired by alternating between a single full MS scan followed by three tandem MS scans on the three most intense precursor masses (as determined by Xcalibur software in real time) from the full scan. The collision energy was normalized to 35%. Dynamic mass exclusion windows were set at 2 min, and all of the spectra were measured with an overall mass/charge (m/z) ratio range of 300-1700. Using TurboSEQUEST (Bioworks Browser 3.1; ThermoElectron), we created an Avian nrpd subset exactly as described.10 Our experimental mass spectra and tandem mass spectra were searched against a combined database containing all annotated chicken and MDV proteins (including cystein carbamidomethylation (Delta mass 57.02 Da) and methionine mono- and dioxidation (Delta mass 16 and 32 Da). The peptide (MS precursor ion) mass tolerance was 1.5 Da and the fragment ion (MS2) tolerance was 1.0 Da. Protein identifications made with peptides that had Xcorr g1.5, 2.0, and 2.5 for +1, +2, and +3 charged ions, respectively, and delta Cn values of g0.118,19 were used for protein identification. Furthermore, the protein identifications were considered valid only when peptides were g6 amino acids long.17 Differential Proteomics Analysis Using PathwayStudio. Proteomics analysis was done using a commercially available software PathwayStudio, Ariadne Genomics, Inc. We imported the chicken database from the Ariadne Genomics download center (http://www.ariadnegenomics.com/downloads/ ?page)pathway&group)support#groups) into PathwayStudio. The chicken database contains protein functional annotations includingGOandvariouspathwaysandnetworksofprotein-protein interactions. We next uploaded the protein lists from the naïve, SEB-activated CD4+ and MD-transformed cells into PathwayStudio. We used the “find pathways” functionality to search for all the GO signaling pathways associated with the data sets. The results ”find pathways” depend on what collection of pathways are present in the database. PathwayStudio calculates the overlap between the protein list/genes and the signaling pathways in the database using the Fisher exact test and assigns a p-value which shows the probability of being associated with a particular pathway among all pathway entities in the database by chance alone.8 We sorted the pathways based on p-value and selected those with p < 0.001 for further comparative analysis. PathwayStudio provides tools for comparing two or
more data sets both at single protein level or protein-protein interaction level (pathways or functions). This makes it possible to find Pathways or functions that are shared or are unique in data sets being compared. For example, selecting Edit f Combine Pathways f Union option leads to formation of a new data set containing a combination of all entities from two or more parent pathway lists. However, when the Intersect option is selected, a new list is formed that contain entities that are shared by the parent data sets, and when the Subtract option is selected, a new list is formed that contain entities that are present only in one parent data set but not the other (differentially expressed pathways). Using these functionalities, we obtained pathways that are unique, shared, or differentially expressed in naïve CD4+ T cells, SEB-activated CD4+ T cells and the CD4+ MD-transformed T cell lymphoma cell line.
Results and Discussion The naïve CD4+- and SEB-activated CD4+ -T cells that were isolated from PBMC using magnetic cell sorting were 94.5 ( 0.5% CD4+. The average number of unique proteins identified from triplicate samples of naïve CD4+, SEB-activated CD4+ and MD-transformed CD4+ cells were 802 ( 160, 1491 ( 286, and 2007 ( 412, respectively. Three protein extracts from same sample were subjected to tandem mass spectrometry, and unique proteins that were identified were pooled for eventual pathways analysis as described.10 We extracted proteins 3 times from the same sample in order to increase the number of proteins identified. Identification of proteins by 2-D LC tandem mass spectrometry has a random component, though is not 100% random. Some proteins are always identified (those present in the highest concentrations); some are often identified and some rarely. This means first that there is not a 100% overlap in proteins identified between each replicate; second not all proteins will be identified in a single run and the more times a sample is replicated, the more proteins will be identified.2 It was previously demonstrated that the first 3 replicates can lead to as much as 80% of analytical completeness, but between 7 and 10 replicates may be needed to equal or exceed a 95% analytical completeness.20 We uploaded the proteins into PathwayStudio for analysis. The number of proteins that were found in the PathwayStudio database represented 90.54% for naïve, 90.67% for SEBactivated, and 89.99% for the MD-transformed CD4+ T cells, which is sufficient to allow analysis of signaling pathways. We used the tools available in PathwayStudio to identify signaling pathways that were expressed in the 3 samples. We selected pathways that were expressed at p < 0.001 level of significance for further analysis and we identified 30 pathways from naïve CD4+ (Supplementary Table 1), 54 pathways from SEBactivated CD4+ (Supplementary Table 2), and 242 pathways from the MD-transformed CD4+ cells (Supplementary Table 3). We used Pathwaystudio to compare and contrast the signaling pathways (p < 0.001) from the samples. Pathways that were differentially expressed due to SEB-activation and MDtransformation are shown in Supplementary Tables 4 and 5, respectively. We expressed the signaling pathways in percentages (%) to give an impression of their relative abundance. For example, in Table 1, pathways that were differentially expressed in SEB-activated CD4+ T cells as compared to naïve CD4+ T cells were 51 (Supplementary Table 4), but only one pathway involved CD40L signaling. The proportion of the CD40L signaling pathway to the total number of pathways is therefore 1/51 or 1.96%. However, in cases where two or more pathways Journal of Proteome Research • Vol. 7, No. 6, 2008 2383
research articles
Buza and Burgess
Table 4. Summary of Signaling Pathways That Are Unique for Marek’s Disease Virus-Transformed Chicken CD4+ T Cell Lymphoma Cell Line As Compared to Naïve Chicken Peripheral Blood CD4+ T Cells and Staphylococcal Enterotoxin B-Activated Chicken Peripheral Blood CD4+ T Cells category
Soluble factors/ Soluble factor receptor mediated
subcategory
Cytokines
Hormones
Nonsoluble factor mediated
signaling pathway/pathway group
5.97 5.97 3.98 3.98 3.48 1.99 1.49 1.00 1.00 1.00 1.00 1.00 5.97 2.49 1.99 1.99 1.49 1.99 1.49 1.00 1.00 0.50 0.50 1.49 1.99 1.00 1.00 1.00 0.50
Fos
5.97
Proteinase-activated receptor 1 precursor (F2R) Notch Activin Receptor (ACVR) Cholecystokinin receptor (CCKR) Endothelin receptor (EDNRB) Integrin signaling (ITGAM/ITGB Cannabinoid receptor (CNR) Dopamine receptor (DRD2) Stromal cell-derived factor 1 (SDF1) f Interleukin 6 signal transducer (IL6ST) Jnk-mapk Kit ligand/Mast cell growth factor/Stem cell factor (KITLG) f Mast/ stem cell growth factor receptor Leptin receptor (LEPR) f CCAAT/enhancer binding protein (CEBPB) Melanocortin 1 receptor (MC1R) f Mothers against decapentaplegic homologue 4 (MADH4) Plasminogen (PLG)-Low density lipoprotein-related protein 1 (LRP1) Prolactin (PRL)-prolactin receptor (PRLR) Stromal cell-derived factor 1 (SDF1) f CD4 Selectin P ligand (SELPLG) f intercellular adhesion molecule 1 (CAM1) Tyrosine kinase, endothelial (TEK) f CCAAT/enhancer binding protein (C/EBP), beta (CEBPB) thrombospondin 1 (THBS1)-CD36 Triple functional domain (TRIO) f Serum response factor (SRF) Anaplastic lymphoma kinase (ALK) f Signal transducer and activator of transcription 3 (STAT3) CD38 f Transcription factor 3 (TCF3) CD4 f Signal transducer and activator of transcription 6 (STAT6) Cholinergic receptor, nicotinic, alpha 7 (CHRNA7) f CAMP responsive element binding protein 1 (CREB1) Ephrin type-B receptor 3 (EPHB3) Acetylcholine receptor (ACH-R) in apoptosis Angiotensinogen (AGT)-Low density lipoprotein-related protein 2 (LRP2)
3.48 3.48 2.49 1.99 1.99 1.99 1.00 1.00 0.50
Journal of Proteome Research • Vol. 7, No. 6, 2008
NOTCH1 NOTCH1 NOTCH1 NOTCH1
% subcategory % category
7.96
Fibroblast growth factor/FGFR MET/Hepatocyte growth factor (HGF) Epidermal growth factor/EGFR Nerve growth factor/NGFR Vascular endothelial growth factor (VEGF) Insulin like growth factor (IGF) Leukemia inhibotory factor (LIF) Brain-derived neurotrophic factor precursor (BDNF) Colony stimulating factor macrophage (CSF1) Neuregulin 1 Neurotrophin (NTF3) Platelet-derived growth factor (PDGF) Tumor necrosis factor receptor Interleukin 6/IL-6R Inteferon beta/gamma/IFNR Inhibin beta A (INHBA) chain Oncostatin M (OSM) Interleukin 4/IL-4R Interleukin 1/IL-1R Interleukin IL-2R Secreted phosphoprotein 1 (SPP1) Interleukin 11 Interleukin 12 C-C Cemokine receptor type 2 (MCP-1R) Insulin (INS) Anti-Muellerian hormone type-2 receptor Gonadotropin-releasing hormone receptor Prostaglandin F (PGF) Erythropoietin
shared common signaling, we summarized the results by combining these pathways into one pathway group. For example, the following seven (7) pathways shared NOTCH1 signaling (Supplementary Table 2). 2384
%
Growth factors Tranforming growth factor B/TGFB receptor
f f f f
0.50 0.50 0.50 0.50 0.50 0.50 0.50 0.50 0.50 0.50 0.50 0.50 0.50 0.50 0.50 0.50 0.50 0.50
MYOD1 signaling pathway. MYT1 signaling pathway. LEF1 signaling pathway. FOS signaling pathway.
39.80
67.16
21.89
5.47
32.84
Different Signaling Pathways Table 5. Signaling Pathways That Are Unique for Naïve Chicken Peripheral Blood CD4+ T Cells As Compared to Staphylococcual Enterotoxin B-Activated Chicken Peripheral Blood CD4+ T Cells and Marek’s Disease Virus-Transformed Chicken CD4+ T Cell Lymphoma Cell Line Signaling Pathway
Activin receptor type 2B (AcvR2B) f Bone morphogenetic protein receptor type IB (BMPR1B) Neural cell adhesion molecule 1 (NCAM1) f Signal transducer and activator of transcription 3 (STAT3) Neurotrophic tyrosine kinase, receptor, type 2 (NTRK2) f Early growth response 1 (EGR1) Neurotrophic tyrosine kinase, receptor, type 3 (NTRK3) f POU domain, class 4, transcription factor 1 (POU4F1) Toll-like receptor 2 (TLR2) f High-mobility group box 1 (HMG1) Insulin-like growth factor 2 receptor (IGF2R) f E74-like factor 4 (ELF4) v-erb-b2 erythroblastic leukemia viral oncogene homologue 3 (ERBB3) f Nuclear factor NF-kappa-B1 (NFKB1) Glycolysis/Gluconeogenesis Table 6. Signaling Pathways That Are Shared between Staphylococcal Enterotoxin B-Activated Chicken Peripheral Blood CD4+ T Cells and Marek’s Disease Virus-Transformed Chicken CD4+ T Cell Lymphoma Cell Line signaling pathways
Integrin Signaling integrin alpha9beta1 integrin alpha6beta1 Tumor necrosis factor receptor superfamily, member 10b (TNFRSF10B) f Rel proto-oncogene protein (REL) Notch homologue 1, translocation-associated (NOTCH1) f Myogenic differentiation 1 (MYOD1) Notch homologue 1, translocation-associated (NOTCH1) f Myelin transcription factor 1 (MYT1) Notch homologue 1, translocation-associated (NOTCH1) f Lymphoid enhancer-binding factor 1 (LEF1) Notch homologue 1, translocation-associated (NOTCH1) f v-fos FBJ murine osteosarcoma viral oncogene homologue (FOS) Notch homologue 1, translocation-associated (NOTCH1) f Signal transducer and activator of transcription 3 (STAT3) Notch homologue 1, translocation-associated 1 (NOTCH1) f Myocyte enhancer factor 2C (MEF2C) Inducible T-cell co-stimulator ligand (B7H2) f CD28 Delta-like protein 1 (DLL1) f NOTCH1 Hepatocyte growth factor (HGF) f CD44 Hepatocyte growth factor (HGF) f Hepatocyte growth factor receptor (MET) Interferon, alpha 1 (IFNA1) f Complement component (3d/Epstein-Barr virus) receptor 2 (CR2) Interferon, alpha 1 (IFN1) f Interferon (alpha, beta and omega) receptor 1(IFNAR1) Interferon, alpha 1 (IFN1) f Interferon (alpha, beta and omega) receptor 2 (IFNAR2) Interferon gamma (IFN G) f Interferon-gamma receptor alpha (IFNGR1) Transforming growth factor, beta 1 (TGFB 1) f CD44
NOTCH1 f STAT3 signaling pathway. NOTCH1 f HEY1 signaling pathway. NOTCH1 f MEF2C signaling pathway. The seven NOTCH1 signaling pathways were combined into one pathway group named NOTCH signaling. The proportion of NOTCH1 signaling pathway group to the total number of pathways is 7/51 or 13.73%.
research articles Flow cytometry analysis indicated that SEB-activated CD4+ T cells were indeed activated as monitored by expression of MHC class II and blast transformation as previously reported.11,21 The CD4+ populations in the activated cells contained higher (p < 0.0001) proportion of MHC-IIhigh CD4+ T lymphocytes (39 ( 0.12%) as compared to unstimulated cells (13 ( 0.11%). Similarly SEB-activated cells contained higher (p < 0.0001) proportion of blasts (22.48 ( 12%) as compared to unstimulated cells (4.95 ( 0.04%) which indicates activation-induced cell proliferation. Pathway analysis information confirmed the activation state of SEB-activated CD4+ T cells. A summary of signaling pathways that were differentially expressed in SEB activated CD4+ T cells are shown in Table 1. Pathways that were differentially expressed due to SEB activation were mainly associated with cell activation, proliferation, inflammation, and cell death. These changes are consistent with literature on SEB action on PBMC and in vivo.16 SEB activates cells by engaging the MHC class II molecules on monocyte/macrophages and Vβ regions of T cell receptors resulting in activation of both monocytes/macrophages, cytokine production, expression of adhesion molecules and or T cell proliferation.16 For effective SEB activation of T cells, however, a costimulation signal involving B7H2 on monocytes/macrophages and CD28 on T cells is essential for induction of IL-2 and subsequent cell proliferation.22 Both the BHB7-CD28 and IL-2 signaling were significantly expressed indicating that the cells were in state of activation and proliferation. Consistent with T cell activation is the increased integrin signaling and cytokine signaling pathways that were enhanced. Cytokine IL-2 and the inflammatory cytokine IFN-γ were increased in SEB-activated human CD4+ and CD8+ T cells23 similar to observation from this study. Stimulation of the T cell through the T-cell receptor causes activation-induced cell death and involves CD9524 or TNF-R1.25,26 Furthermore, the fate of SEB-exposed T cells is cell-death or anergy.16,27 Our results agree and demonstrate that death receptor signaling, including TNFR and caspases, was significantly expressed in SEB-activated cells. Notch is an essential signaling receptor that contributes to proper cellular development and also influences cell fate, proliferation, and survival.28 Similar to our results, Notch signaling was enhanced in SEBexposed mouse CD4+ T cells.27 The direct function of Notch signaling in SEB-activated cells is not yet determined but may contribute to decision to destine the activated cells to eventual cell death. A total of 220 pathways were differentially expressed in MD lymphoma transformed CD4+ T cell line as compared to the naïve CD4+ T cells (Supplementary Table 5). We summarized the 220 signaling pathways by classifying them into two categories: the soluble factor/soluble factor receptor mediated category and the nonsoluble factor mediated category, comprising 138 pathways (62.73%) and 82 pathways (37.27%), respectively (Table 2). We further subdivided the 138 soluble factor/soluble factor receptor mediated category into three subcategories, namely, growth factors (37.73%), cytokines (20%), and hormones (55%). Pathways that shared common signaling were combined into one pathway group, and their relative proportion (%) was calculated as described previously. Signaling pathways involving soluble factors, particularly the cell growth factors, predominated. We postulate that these pathways may be stimulated by soluble factors secreted by the lymphoma cell themselves because these soluble factors were present in the proteome of the MD-transformed cells both in this study and our previous work.10 Cancer cell autocrine Journal of Proteome Research • Vol. 7, No. 6, 2008 2385
research articles activation by soluble factors is one mechanism by which cancer cells propagate themselves.29,30 However, our work does not rule out other soluble factor-mediated signaling pathways may be activated by the transformation process itself in vivo because the MD-transformed cells are known to express an activated phenotype.13 Growth factors play multiple and overlapping roles on tumor cells including mitogenesis, motogenesis, antiapoptosis, and angiogenesis.31 The growth factor signaling pathways we observed to be expressed in MD-transformed lymphomas are consistent with malignant transformation, fast growth, and spread of MD-lymphomas. Some growth factor signaling pathways present are associated with oncogenic transformation such as the EGFR, PDGFR, VEGFR, and MET.31 MET, VEGF, Leukemia inhibitory factor, and FGF signaling reported to enhance cancer cell motility which contributes to invasiveness.32 VEGF, MET, and BDGF signaling promote formation of newer blood vessels (angiogenesis) which help cancer cell gain access to nutrients and also migrate to other areas (metastasis).33–35 Notably, NGF signaling is suggested to be important in cancer cell perineural invasion and the spread of cancer cells along nerves.36 Neural invasion is characteristic of MD lymphomas and is responsible for classical signs of the MD which includes lameness, leg paralysis, and drooping of the wings.37 Cytokine signaling pathways that promote the antitumor immune responses and those that promote tumor growth and invasion were significantly expressed in MD lymphoma cells.38,39 Pathways that promote immune tumor elimination included IFNR, IFNβ, IFNγ, IL-12, and TNFR signaling.38,39 Signaling that promote tumor growth by suppressing antitumor immune responses included VEGF and TGFβ.39,40 Although TGF-β may be tumor suppressive,41 it can also play a tumor-promoting/ pro-invasive role depending on presence of other factors such as Ras.42 Furthermore, TGF-β is important in T-regulatory cell development and the fact that we also found IL-12 protein, which is mainly produced by T-regulatory rather than normal T-helper cells, agrees with our reports that the MD-transformed lymphoma cell line10 or the ex vivo-derived MD-transformed lymphoma cells43 have a T-regulatory phenotype. We identified other cytokine, chemokine, and hormone-mediated signaling pathways differentially expressed by the MD-transformed cells, and we have described their possible function in lymphomagenesis previously.10 Signaling pathways in our nonsoluble factor category (Table 2) included those associated with cell-extracellular matrix contact (ITG), cell-cell contact (Notch), cell-surface receptors (proteinase-activated receptor 1/F2R), and intracellular signaling (transcription factor c-fos). In addition to their normal physiological functions, some of these promote cancer. Integrin-mediate cell-matrix contact provides critical signaling that regulates cellular proliferation, migration, and apoptosis.44 Integrin signaling was also enhanced in our previous studies using the MD-transformed lymphoma cell line and was suggested to promote metastasis.10 Members of the Fos signaling are activated by a variety of stimuli, including growth factors, cytokines, neurotransmitters, polypeptide hormones, stress, and cell injury and induce responses that include proliferation and differentiation.45,46 The enhanced soluble factor signaling observed for the MD-transformed proteome may contribute to the enhanced fos-signaling. However, fos is also a protooncogene47 and may contribute to transformation directly. Notch signaling regulates cell-fate decisions including the implementation of differentiation, proliferation, and apoptotic 2386
Journal of Proteome Research • Vol. 7, No. 6, 2008
Buza and Burgess programs. Notch is also suggested to be a proto-oncogene and a mutated form of the gene is associated with development of mouse mammary tumors.48 Proteinase-activated receptor 1/F2R signaling is important in blood coagulation and cell motility but also promotes angiogenesis and has been implicated in prostate tumorigenesis and metastasis.44 Endothelin maintains a delicate balance between vasoconstriction and vasodilation but has also been linked with the induction of survival signals that control resistance to apoptosis.49 We also used functionalities in Pathwaystudio to find pathways that were unique (only found in one sample but absent in the other two samples). Pathways that were unique for SEB-activated CD4+ T cells represented those associated with inflammation and cell death (Table 3) while pathways unique for MD-transformed cells reflected a strong soluble factor response (Table 4). Pathways that were unique for naïve CD4+ T cells included gluconeogenesis (Table 5), suggesting that the SEB-activated and MD-transformed cells switch to other energy production pathways to satisfy metabolic requirements. Pathways that were common between SEB-activated and MD-transformed cells were associated with activation as expressed by cell-intercellular matrix interaction (integrin signaling), cell fate decisions (Noth signaling), and immune response (interferon signaling) (Table 6). In conclusion, we used the differential proteomics functionalities provided by Pathwaystudio to compare and contrast the signaling pathways expressed in proteomes of chicken CD4+ T cells that were exposed to different stimuli. The results obtained concur with the available literature and also provide new information on cellular changes induced by SEB as well as MDV-transformation.
Acknowledgment. This work was supported by a USDA NRI 2004-35204- 14829. We acknowledge Tibor Pechan for running the mass spectrometer. We acknowledge the excellent suggestions made by an expert, but anonymous reviewer. This paper is Mississippi Agricultural and Forestry Experimental Station publication number J-11291. Supporting Information Available: Supplementary tables showing signaling pathways identified from the three samples. This material is available free of charge via the Internet at http://pubs.acs.org. References (1) Gygi, S. P.; Corthals, G. L.; Zhang, Y.; Rochon, Y.; Aebersold, R. Evaluation of two-dimensional gel electrophoresis-based proteome analysis technology. Proc. Natl. Acad. Sci. U.S.A. 2000, 97 (17), 9390–5. (2) Baggerman, G.; Vierstraete, E.; De Loof, A.; Schoofs, L. Gel-based versus gel-free proteomics: a review. Comb. Chem. High Throughput Screening 2005, 8 (8), 669–77. (3) Monteoliva, L.; Albar, J. P. Differential proteomics: an overview of gel and non-gel based approaches. Briefings Funct. Genomics Proteomics 2004, 3 (3), 220–39. (4) Tan, F. K.; Hildebrand, B. A.; Lester, M. S.; Stivers, D. N.; Pounds, S.; Zhou, X.; Wallis, D. D.; Milewicz, D. M.; Reveille, J. D.; Mayes, M. D.; Jin, L.; Arnett, F. C., Jr. Classification analysis of the transcriptosome of nonlesional cultured dermal fibroblasts from systemic sclerosis patients with early disease. Arthritis Rheum. 2005, 52 (3), 865–76. (5) Zhou, X.; Tan, F. K.; Xiong, M.; Arnett, F. C.; Feghali-Bostwick, C. A. Monozygotic twins clinically discordant for scleroderma show concordance for fibroblast gene expression profiles. Arthritis Rheum. 2005, 52 (10), 3305–14. (6) Tan, F. K.; Zhou, X.; Mayes, M. D.; Gourh, P.; Guo, X.; Marcum, C.; Jin, L.; Arnett, F. C., Jr. Signatures of differentially regulated interferon gene expression and vasculotrophism in the peripheral
research articles
Different Signaling Pathways
(7) (8) (9)
(10) (11)
(12)
(13) (14) (15) (16) (17)
(18) (19)
(20)
(21)
(22)
(23) (24) (25) (26)
blood cells of systemic sclerosis patients. Rheumatology (Oxford) 2006, 45 (6), 694–702. Attwood, T. K. Genomics. The Babel of bioinformatics. Science 2000, 290 (5491), 471–3. Yuryev, A.; Mulyukov, Z.; Kotelnikova, E.; Maslov, S.; Egorov, S.; Nikitin, A.; Daraselia, N.; Mazo, I. Automatic pathway building in biological association networks. BMC Bioinf. 2006, 7, 171. Wang, J.; Robinson, J. F.; O’Neil, C. H.; Edwards, J. Y.; Williams, C. M.; Huff, M. W.; Pickering, J. G.; Hegele, R. A. Ankyrin G overexpression in Hutchinson-Gilford progeria syndrome fibroblasts identified through biological filtering of expression profiles. J. Hum. Genet. 2006, 51 (11), 934–42. Buza, J. J.; Burgess, S. C. Modeling the proteome of a Marek’s disease transformed cell line: a natural animal model for CD30 overexpressing lymphomas. Proteomics 2007, 7 (8), 1316–26. Burgess, S. C.; Davison, T. F. Identification of the neoplastically transformed cells in Marek’s disease herpesvirus-induced lymphomas: recognition by the monoclonal antibody AV37. J. Virol. 2002, 76 (14), 7276–92. Burgess, S. C.; Young, J. R.; Baaten, B. J.; Hunt, L.; Ross, L. N.; Parcells, M. S.; Kumar, P. M.; Tregaskes, C. A.; Lee, L. F.; Davison, T. F. Marek’s disease is a natural model for lymphomas overexpressing Hodgkin’s disease antigen (CD30). Proc. Natl. Acad. Sci. U.S.A. 2004, 101 (38), 13879–84. Dienglewicz, R. L.; Parcells, M. S. Establishment of a lymphoblastoid cell line using a mutant MDV containing a green fluorescent protein expression cassette. Acta Virol. 1999, 43 (2-3), 106–12. McCormick, J. K.; Yarwood, J. M.; Schlievert, P. M. Toxic shock syndrome and bacterial superantigens: an update. Annu. Rev. Microbiol. 2001, 55, 77–104. Chatila, T.; Geha, R. S. Signal transduction by microbial superantigens via MHC class II molecules. Immunol. Rev. 1993, 131, 43– 59. Krakauer, T. Chemotherapeutics targeting immune activation by staphylococcal superantigens. Med. Sci. Monit. 2005, 11 (9), RA290-5. McCarthy, F. M.; Burgess, S. C.; van den Berg, B. H.; Koter, M. D.; Pharr, G. T. Differential detergent fractionation for non-electrophoretic eukaryote cell proteomics. J. Proteome Res. 2005, 4 (2), 316–24. MacCoss, M. J.; Wu, C. C.; Yates, J. R., III. Probability-based validation of protein identifications using a modified SEQUEST algorithm. Anal. Chem. 2002, 74 (21), 5593–9. Yates, J. R., III; Eng, J. K.; McCormack, A. L. Mining genomes: correlating tandem mass spectra of modified and unmodified peptides to sequences in nucleotide databases. Anal. Chem. 1995, 67 (18), 3202–10. Durr, E.; Yu, J.; Krasinska, K. M.; Carver, L. A.; Yates, J. R.; Testa, J. E.; Oh, P.; Schnitzer, J. E. Direct proteomic mapping of the lung microvascular endothelial cell surface in vivo and in cell culture. Nat. Biotechnol. 2004, 22 (8), 985–92. Mannie, M. D.; Dawkins, J. G.; Walker, M. R.; Clayson, B. A.; Patel, D. M. MHC class II biosynthesis by activated rat CD4+ T cells: development of repression in vitro and modulation by APC-derived signals. Cell Immunol. 2004, 230 (1), 33–43. Fraser, J. D.; Newton, M. E.; Weiss, A. CD28 and T cell antigen receptor signal transduction coordinately regulate interleukin 2 gene expression in response to superantigen stimulation. J Exp Med 1992, 175 (4), 1131–4. Liu, J.; Wu, C. Y. [Subpopulations and cytokine expression of naive and memory T cells in normal human PBMCs.]. Xibao Yu Fenzi Mianyixue Zazhi 2007, 23 (1), 2–5. Krueger, A.; Fas, S. C.; Baumann, S.; Krammer, P. H. The role of CD95 in the regulation of peripheral T-cell apoptosis. Immunol. Rev. 2003, 193, 58–69. Green, D. R.; Droin, N.; Pinkoski, M. Activation-induced cell death in T cells. Immunol. Rev. 2003, 193, 70–81. Arnold, R.; Brenner, D.; Becker, M.; Frey, C. R.; Krammer, P. H. How T lymphocytes switch between life and death. Eur. J. Immunol. 2006, 36 (7), 1654–8.
(27) Kurella, S.; Yaciuk, J. C.; Dozmorov, I.; Frank, M. B.; Centola, M.; Farris, A. D. Transcriptional modulation of TCR, Notch and Wnt signaling pathways in SEB-anergized CD4+ T cells. Genes Immun. 2005, 6 (7), 596–608. (28) Baron, M. An overview of the Notch signalling pathway. Semin. Cell Dev. Biol. 2003, 14 (2), 113–9. (29) Goustin, A. S.; Leof, E. B.; Shipley, G. D.; Moses, H. L. Growth factors and cancer. Cancer Res. 1986, 46 (3), 1015–29. (30) Sporn, M. B.; Roberts, A. B. Autocrine growth factors and cancer. Nature 1985, 313 (6005), 745–7. (31) Breuhahn, K.; Longerich, T.; Schirmacher, P. Dysregulation of growth factor signaling in human hepatocellular carcinoma. Oncogene 2006, 25 (27), 3787–800. (32) Sun, S.; Schiller, J. H. Angiogenesis inhibitors in the treatment of lung cancer. Crit. Rev. Oncol. Hematol. 2007, 62 (2), 93–104. (33) Zetter, B. R. Cell motility in angiogenesis and tumor metastasis. Cancer Invest. 1990, 8 (6), 669–71. (34) Sooriakumaran, P.; Kaba, R. Angiogenesis and the tumour hypoxia response in prostate cancer: a review. Int. J. Surg. 2005, 3 (1), 61– 7. (35) Kermani, P.; Hempstead, B. Brain-derived neurotrophic factor: a newly described mediator of angiogenesis. Trends Cardiovasc. Med. 2007, 17 (4), 140–3. (36) Malcangio, M.; Garrett, N. E.; Cruwys, S.; Tomlinson, D. R. Nerve growth factor- and neurotrophin-3-induced changes in nociceptive threshold and the release of substance P from the rat isolated spinal cord. J. Neurosci. 1997, 17 (21), 8459–67. (37) Payne, L. N.; Frazier, J. A.; Powell, P. C. Pathogenesis of Marek’s disease. Int. Rev. Exp. Pathol. 1976, 16, 59–154. (38) Gresser, I.; Belardelli, F. Endogenous type I interferons as a defense against tumors. Cytokine Growth Factor Rev. 2002, 13 (2), 111–8. (39) Smyth, M. J.; Cretney, E.; Kershaw, M. H.; Hayakawa, Y. Cytokines in cancer immunity and immunotherapy. Immunol. Rev. 2004, 202, 275–93. (40) Ohm, J. E.; Carbone, D. P. VEGF as a mediator of tumor-associated immunodeficiency. Immunol. Res. 2001, 23 (2-3), 263–72. (41) Siegel, P. M.; Massague, J. Cytostatic and apoptotic actions of TGFbeta in homeostasis and cancer. Nat. Rev. Cancer 2003, 3 (11), 807–21. (42) Gotzmann, J.; Huber, H.; Thallinger, C.; Wolschek, M.; Jansen, B.; Schulte-Hermann, R.; Beug, H.; Mikulits, W. Hepatocytes convert to a fibroblastoid phenotype through the cooperation of TGF-beta1 and Ha-Ras: steps towards invasiveness. J. Cell Sci. 2002, 115 (Pt 6), 1189–202. (43) Shack, A. L.; Buza, J. J.; Burgess, S. C., The neoplastically transformed (CD30hi) Marek’s disease lymphoma cell phenotype most closely resembles T-regulatory cells. Cancer Immunol. Immunother. 2008, in press. (44) Weyant, M. J.; Carothers, A. M.; Bertagnolli, M. E.; Bertagnolli, M. M. Colon cancer chemopreventive drugs modulate integrinmediated signaling pathways. Clin. Cancer Res. 2000, 6 (3), 949– 56. (45) Ashcom, G.; Gurland, G.; Schwartz, J. Growth hormone synergizes with serum growth factors in inducing c-fos transcription in 3T3F442A cells. Endocrinology 1992, 131 (4), 1915–21. (46) Milde-Langosch, K. The Fos family of transcription factors and their role in tumourigenesis. Eur. J. Cancer 2005, 41 (16), 2449– 61. (47) Sassone-Corsi, P. Signaling pathways and c-fos transcriptional responses--links to inherited diseases. N. Engl. J. Med. 1995, 332 (23), 1576–7. (48) Dontu, G.; Jackson, K. W.; McNicholas, E.; Kawamura, M. J.; Abdallah, W. M.; Wicha, M. S. Role of Notch signaling in cell-fate determination of human mammary stem/progenitor cells. Breast Cancer Res. 2004, 6 (6), R605-15. (49) Peduto Eberl, L.; Bovey, R.; Juillerat-Jeanneret, L. Endothelinreceptor antagonists are proapoptotic and antiproliferative in human colon cancer cells. Br. J. Cancer 2003, 88 (5), 788–95.
PR700844Z
Journal of Proteome Research • Vol. 7, No. 6, 2008 2387
