Nuclear Proteome Dynamics in Differentiating Embryonic Carcinoma

May 11, 2010 - (2) The differentiation of the well-established embryonic carcinoma (EC) cell line, NTERA-2, is a robust, reproducible and controllable...
1 downloads 0 Views 4MB Size
Download PDF
Nuclear Proteome Dynamics in Differentiating Embryonic Carcinoma (NTERA-2) Cells Emma Pewsey,† Christine Bruce,† Peter Tonge,‡ Caroline Evans,§ Saw Yen Ow,§ A. Stephen Georgiou,| Phillip C. Wright,§ Peter W. Andrews,‡ and Alireza Fazeli*,† Academic Unit of Reproductive and Developmental Medicine, University of Sheffield, Level 4, The Jessop Wing, S10 2SF Sheffield, United Kingdom, The Centre for Stem Cell Biology, Department of Biomedical Science, University of Sheffield, Western Bank, Sheffield, S10 2TN, United Kingdom, Department of Chemical and Process Engineering, ChELSI Institute, Mappin Street, Sheffield, S1 3JD, United Kingdom, and Centre for Developmental Genetics, School of Medicine and Biomedical Science, University of Sheffield, Firth Court, Western Bank, Sheffield S10 2TN, United Kingdom Received November 22, 2009

The use of stem cells for generating cell types suitable for therapy is dependent on understanding the mechanisms, and identifying biomarkers, that control cell fate into different lineages. In this study, we aimed to characterize the nuclear protein dynamics of NTERA-2 cells undergoing retinoic acid-induced differentiation. We focused specifically on the first six days of differentiation, to provide insight into the earliest differentiation events, and employed techniques to specifically monitor the nuclear proteome. Well-characterized gene expression markers were used to precisely stage cell differentiation across the experimental time course. A combination of the novel iTRAQ and ExacTag labeling technologies, together with LC-ESI tandem mass spectrometry, were then used to accurately measure nuclear protein expression changes occurring within these differentiation-staged cells. We report proteins that showed significantly altered expression over the first 6 days of differentiation. Extensive bioinformatic analysis was undertaken, resulting in the construction of a novel interactome network, which revealed the temporal dynamics of the nuclear protein network in the context of neuronal differentiation. Keywords: neuronal differentiation • embryonic carcinoma • ExacTag • iTRAQ • human protein-protein interactome • nuclear protein dynamics

Introduction An important goal within the field of stem cell research is to understand how stem cells can proliferate extensively and give rise to a multitude of specific cell lineages. Obtaining pure populations of specific phenotypes from embryonic stem (ES) cells may facilitate the generation of isogenic replacement cells for the treatment of a variety of diseases.1 However, the use of human ES cells for generating cell types suitable for therapy is dependent on understanding the mechanisms and identifying the biomarkers that control cell fate into the different lineages. Neuronal differentiation is a well-established system of differentiation. Currently, there are a number of successful methods available to produce neurons, astrocytes and immature oligodendrocytes from mitotically active progenitor cells in a manner similar to fetal development.2 The differentiation of the well-established embryonic carcinoma (EC) cell line, * To whom correspondence should be addressed. Dr Alireza Fazeli, Academic Unit of Reproductive and Developmental Medicine, University of Sheffield, Level 4, The Jessop Wing, S10 2SF Sheffield, United Kingdom. Tel: +44 114 2268195. Fax: +44 114 2268538. E-mail: [email protected]. † Academic Unit of Reproductive and Developmental Medicine, University of Sheffield. ‡ Department of Biomedical Science, University of Sheffield. § ChELSI Institute. | School of Medicine and Biomedical Science, University of Sheffield.

3412 Journal of Proteome Research 2010, 9, 3412–3426 Published on Web 05/11/2010

NTERA-2, is a robust, reproducible and controllable system that offers a simple model of differentiation. The NTERA-2 cell line is pluripotent and differentiates into various cell types when exposed to certain stimuli;3-5 this makes it a convenient model pertinent to human development. When cultured in the presence of retinoic acid (RA), NTERA-2 cells differentiate into a variety of cell types including the welldeveloped postmitotic central nervous system neurons.3,6 NTERA-2 derived neurons have been well studied. For example, they express all three neurofilament proteins7 and N-methyl8 9 D-aspartate (NMDA). They also establish functional synapses. Consequently, there is plenty of information about the structure and function of the terminally differentiated neurons. However, little is known about the earlier mechanisms that lead to the development of these neurons and the commitment of these undifferentiated NTERA-2 cells to the neural lineage. Most research to date has focused on studying one or a few genes involved in neurogenesis. However, the neuronal developmental program represents a complex spatial-temporal expression profile, which is too sophisticated for such single gene analyses. While an attempt has been made to map the transcriptional profile of RA-induced NTERA-2 differentiation using high throughput technologies,2 it is well established that transcript abundance often does not correlate with protein 10.1021/pr901069d

 2010 American Chemical Society

Nuclear Protein Dynamics of Differentiating NTERA-2 Cells 10-12

abundance. In this case, the functional significance of much of the transcriptional regulation remains unknown. To date, there is no information regarding the global proteomic alterations occurring in the early stages of RAinduced NTERA-2 differentiation. Those that have used a proteomic approach for the analysis of neural differentiation in other model systems have shown how effective the approach can be.13 In this study, we used high throughput quantitative MSbased technologies to characterize the nuclear protein dynamics of NTERA-2 cells undergoing RA-induced neuronal differentiation. We focused specifically on the first 6 days of differentiation to provide insight into the very earliest differentiation events. Furthermore, as it is ultimately alterations in nuclear protein expression that direct differentiation, we employed techniques to specifically enrich and monitor the nuclear proteome. Our findings reveal key nuclear proteins and transcription factors involved in controlling the fate of NTERA-2 cells into the neuronal lineage. A novel interactome map was created, which incorporated the temporal dynamics of the nuclear protein network. These findings provide a unique view of the molecular events occurring early on in neuronal differentiation. Additionally, these findings may provide useful insights into the reverse process of differentiation, the phenomenon of nuclear reprogramming,14-16 where identification of master and early regulators of differentiation is required.

Experimental Procedures Cell Culture Conditions. NTERA-2 cells17 were grown in 175 cm2 tissue culture flasks (T175, Corning Costar Europe, High Wycombe, U.K.) and maintained at 37 °C in a humidified atmosphere of 10% CO2. Cells were grown in DMEM growth medium (Dulbecco‘s Modified Eagle‘s Medium, Sigma-Aldrich, Poole, U.K.), and supplemented with 10% fetal calf serum (FCS) (Invitrogen, Paisley, U.K.) and 1% L-glutamine (Invitrogen). Subculture of Cells. Cells were routinely passaged upon reaching confluency by means of mechanical detachment. Depending upon cell proliferation and cell density, NTERA-2 cells were passaged every 3 days by splitting a confluent flask of cells at a ratio of 1:3. Growth medium was aspirated, leaving a small volume of media (∼2 mL) to roll a number of glass beads around. Detached cells were suspended in 5 mL of media and replated in to three new flasks. NTERA-2 cells used for experimentation were always below a cell passage number of 50 to minimize any genotype or phenotype alterations that may occur during prolonged in vitro culture. RA-Mediated Cell Differentiation. A confluent T175 flask of NTERA-2 stem cells was incubated in 2 mL trypsin at 37 °C, until detachment of cells from the bottom of the flask. Trypsin was inactivated by the addition of 20 mL of standard growth medium and cell suspension centrifuged at 300× g for three minutes. After cell number was calculated, cells were plated at 106 cells per 175 cm2 flask in normal growth media supplemented with 10-5 M all-trans-retinoic acid (Sigma). The alltrans-retinoic acid (RA) was stored as a DMSO diluted stock solution of 10-2 M and diluted 1:1000 in growth media. RA supplemented growth media was replenished every three days. NTERA-2 cells were incubated in the presence of retinoic acid for 0, 1, 3, and 6 days. Nuclear Protein Extraction. The nuclear protein extraction protocol was adapted from Desrivieres et al.18 Briefly, cells were centrifuged at 300× g for 5 min at 4 °C and the supernatant discarded. The resulting cell pellet was resuspended in 10

research articles volumes of Buffer A (pH 8.0, 50 mM NaCl, 10 mM HEPES, 0.5 M sucrose,1 mM EDTA, 0.25 mM EGTA, 0.5 mM spermidine, 0.5% triton-X-100 and Protease inhibitors) and vortexed vigorously. After vortexing, the samples were centrifuged at 1000× g for 10 min at 4 °C and the supernatant, containing cytoplasmic proteins, was removed. The resulting cell pellet was resuspended in 1 mL of Buffer B (pH 8.0, 10 mM HEPES, 25% glycerol, 500 mM NaCl, 0.1 mM EDTA, 0.1 mM EGTA, 0.5 mM spermidine, 0.25 mM DTT and Protease inhibitors) and mixed on a rotating wheel at 4 °C for 30 min. Samples were centrifuged at 4000× g for 10 min at 4 °C and the supernatant containing nuclear proteins was collected. The 2D Quant protein assay (GE Healthcare, Amersham, UK) was performed to determine the protein concentration of all samples. Absorbance was read at 480 nm using a Benchmark 96 well plate reader (Bio-Rad, Hemel Hempstead, U.K.). The nuclear protein fractions were stored at -80 °C. The efficacy of this nuclear extraction protocol was first validated using a human HeLa cell line. Once this technique was validated, the same procedure was used on NTERA-2 cells removed from the culture flasks at each time point (0, 1, 3, and 6 days) of differentiation using trypsin (see above) and washed in PBS (Invitrogen). Western Blot Analysis. To verify the enrichment of nuclear proteins, 10 µg of protein from the nuclear and cytoplasmic fractions was separated by SDS-PAGE using 12.5% polyacrylamide gels. Resolved proteins were transferred to a PVDF Immobilon PSQ transfer membrane (0.2-µm pore size) (Millipore, Watford, U.K.) and membranes were blocked with 3% (w/v) nonfat milk powder in Tris-buffered saline containing 0.1% v/v Tween 20 (TTBS) for 2 h at room temperature. Blocked membranes were then incubated overnight at 4 °C with rabbit antiactin (catalogue number: A2006, Sigma) or rabbit antiHistone H3 antibody (catalogue number: AB1791, AbCam, Cambridge, U.K.), diluted at 1:1000 and 1:5000 respectively in 3% (w/v) milk powder in TTBS. Membranes were incubated in a donkey antirabbit horseradish peroxidase-conjugated secondary antibody (catalogue number 111-035-003, Jackson ImmunoResearch Laboratories, Inc.) for 1.5 h (1:2500 dilution). Immunoreactive proteins on the membranes were detected using SuperSignal West Dura chemiluminescent reagents (Perbio) and exposure to Hyperfilm ECL high performance chemiluminescent film (GE Healthcare). Western blotting was also used to confirm the changes in protein expression in NTERA2 cells witnessed by iTRAQ and ExacTag. Blocked membranes were incubated with either of the following antibodies: rabbit anti-Tropomyosin 1 (alpha) polyclonal antibody (catalogue number: AB55915, AbCam, Cambridge, U.K.) or mouse antibeta actin monoclonal antibody (catalogue number: AB8226, AbCam, Cambridge, U.K.). Antibodies were diluted 1:800 and 1:2000, respectively, in 1% (w/v) milk powder in TTBS. RNA Isolation, cDNA Synthesis and Reverse Transcription Polymerase Chain Reaction (RT-PCR). Total RNA was extracted from NTERA-2 cells after 6 days of culture using TRI reagent and following the manufacturer’s guidelines (Sigma, Poole, U.K.). Total RNA was treated with DNase I (DNA-free kit, Ambion, Huntingdon, U.K.) to remove any DNA contamination from samples. First-strand cDNA synthesis was performed using Oligo dT primers (Metabion, Martinsried, Germany) and the Superscript II reverse transcriptase system (Invitrogen). RT-PCR was performed using the prepared cDNA, primers for different markers of pluripotency and neuronal development (Table 1) and Platinum Blue PCR Super Mix (Invitrogen) Journal of Proteome Research • Vol. 9, No. 7, 2010 3413

research articles

Pewsey et al.

Table 1. Sequences of Primers Used in This Study and Melting Temperature, Product Size and Cycle Number for Each Primer Pair primer

forward sequence (5′-3′)

B-actin ATCTGGCACCACACCTTCTACAATGAGCTGCG Oct-4 CGACCATCTGCCGCTTTGAG Sox2 CCCCCGGCGGCAATAGCA Eomesodermin TCACCCCAACAGAGCGAAGAGG NeuroD1 AAGCCATGAACGCAGAGGAGGACT Pax6 GCTGCCAGCAACAGGAAGGAG Hash1 CGCGTGTGCTGCTCCCTTCT Nucleophosmin 1 TTGTTGAAGCAGAGGCAATG Calmodulin 1 TCAGCTGACCGAAGAACAGA Tropomyosin 1 TCATCATTGAGAGCGACCTG Nucleolin GAGGAGGAAGAGCCTGTCAA

using the following conditions: 20-35 cycles in total, 95 °C for 1 min, 58 °C for 30 s and 72 °C for 1 min cycles, with a start time of 95 °C for 3 min and an end time of 4 °C indefinitely. Forward and Reverse primers were either designed using DNAstar software or utilized from papers, and purchased from Metabion (Germany). Primers hybridizing to the mRNA encoding the β-actin cytoskeletal protein were used as an internal standard for RT-PCR. RT-PCR was also used to confirm the changes in protein expression witnessed by iTRAQ and ExacTag. The gene expression of Tropomyosin1, Nuclephosmin 1, Nucleolin, Calmodulin 1 and β-actin were measured (Table 1) using the following conditions: 25 cycles in total, 95 °C for 1 min, 60 °C for 30 s and 72 °C for 1 min cycles, with a start time of 95 °C for 3 min and an end time of 4 °C indefinitely. ExacTag Labeling. Proteins from three biological replicates were pooled to give 100 µg of protein for each sample (Day 0,

reverse sequence (5′-3′)

CGTCATACTCCTGCTTGCTGATCCACATCTGC CCCCCTGTCCCCCATTCCTA TCGGCGCCGGGGAGATACAT AGAGATTTGATGGAAGGGGGTGTC AGCTGTCCATGGTACCGTAA GTGCCCATTGGCTGACTGTTC GGCTCGCCGGTCTCATCCTA TGCATCTTCCTCCACAGCTA GCTTCTGTTGGGTTCTGACC CCTGAGCCTCCAGTGACTTC TGAAAGCCGTAGTCGGTTCT

size cycle Tm (°C) (bp) number

60 60 60 57 55 62 60 60 60 60 60

838 583 448 370 579 600 404 194 132 123 121

24 26 32 33 32 32 34 30 30 30 30

Day 1, Day 3 and Day 6) (Figure 1). A Plus-one 2D clean up kit (GE Healthcare, Amersham, U.K.) was used according to the manufacturer’s instructions to precipitate the protein. The resulting pellets were redissolved in 84 µL of protein dissolution buffer (0.5 M ammonium bicarbonate (NH4HCO3), pH 8.5), proteins were further denatured using 5% SDS. Subsequently the resuspended proteins were reduced in 2 mM tris(2-carboxyethyl)phosphine (TCEP) in accordance with the manufacturers guidelines (PerkinElmer Waltham, MA, USA). Samples were ExacTag labeled in duplicate as follows (100 µg each per label): Day 0, 343; Day 1, 355; Day 3, 406; Day 6, 457; Day 0, 349; Day 1, 400; Day 3, 412; Day 6, 463; BSA standards, 469 and 475. The 10 labeled samples were pooled prior to fractionation by SDS-PAGE. Fractionation by SDS-PAGE. Eight-hundred micrograms of the pooled labeled protein sample was separated by SDS-PAGE using a large 17 cm format 10% polyacrylamide gel. Electro-

Figure 1. Method used for studying the nuclear proteome dynamics in differentiating NTERA-2 cells. NTERA-2 cells were cultured in the presence of retinoic acid for 0, 1, 3, or 6 days before a nuclear protein extraction was performed on the samples. Nuclear protein was isobarically labeled using using iTRAQ or ExacTag and subjected to LC-ESI-MS/MS. 3414

Journal of Proteome Research • Vol. 9, No. 7, 2010

Nuclear Protein Dynamics of Differentiating NTERA-2 Cells phoresis was performed at 20 V overnight, then at 120 V for ∼2 h using a Mini-Protean II gel electrophoresis system (BioRad). The gels were fixed for 30 min in 7% (v/v) acetic acid in 40% (v/v) methanol at room temperature. Gels were then rinsed with water and stained with colloidal brilliant blue Coomassie G50 (Sigma) for 1 h at room temperature and subsequently destained in 10% acetic acid in 25% (v/v) methanol after imaging. The gel lanes containing the labeled protein sample were excised horizontally into 2 mm thick bands, diced further and then destained with 200 mM ammonium bicarbonate with 40% (v/v) acetonitrile (VWR International Ltd., Leicester, U.K.). Proteins were digested with 20 ng/µL of sequencing grade modified trypsin (Promega, Southampton, U.K.) in 50 mM ammonium bicarbonate at 37 °C for 12 h. Peptides were then extracted and transferred into a new siliconised tube. Peptides were extracted in four sequential steps using 30 µL of 25 mM NH4CO3 (10 min, room temperature), 30 µL acetonitrile (15 min, 37 °C), 50 µL of 5% (v/v) formic acid (15 min, 37 °C), and finally, with 30 µL acetonitrile (15 min, 37 °C). All extracts were pooled and dried in a vacuum centrifuge. Mass Spectrometric Analysis - ExacTag. The dried peptides were dissolved in 40 µL 0.1% (v/v) formic acid in 3% (v/v) ACN in water. Samples were centrifuged for 5 min at 12 000× g and then 15 µL of sample was subjected to liquid chromatography electrospray ionization tandem mass spectrometry (LC-ESI-MS/ MS). Mass spectrometry was performed using a Bruker Esquire HCT quadrupole-ion trap tandem mass spectrometer (Bruker Daltonics, Coventry, U.K.), coupled to an online nanocapillary liquid chromatography system (Ultimate 3000, Dionex/LC Packings, Amsterdam, The Netherlands). The peptide mixture was separated on a 15 cm nano-75 PepMap C-18 reverse phase capillary column (LC Packings), with a constant flow rate of 0.3 µL/min. The LC gradient started with 3% buffer B (0.1% formic acid in 97% ACN) and 97% buffer A (0.1% formic acid in 3% ACN) for 3 min, followed by 3-30% buffer B for 90 min, then 90% buffer B for 7 min, and finally 3% buffer B for 8 min. The mass spectrometer was set to perform data acquisition in the positive ion mode, with a selected mass range of 300-1800 m/z up to a set ICC of 200 000 charge and an accumulation time of no lower than 5 Hz. Peptides with +2 and +3 charge states were preferentially selected for tandem mass spectrometry, and the ICC summation was set also at 200 000 charge. The six most abundantly charged peptides greater than the 25 000 count threshold were selected for MS/MS. Ions were then actively excluded for 60 s. iTRAQ Labeling. One hundred micrograms of protein from each sample (Day 0, Day 1, Day 3 and Day 6) was subjected to buffer exchange using Microcon YM-3 Centrifugal filters (Millipore, Watford, U.K.) into 20 µL of 1 M trithylammonium bicarbonate (TEAB). Subsequently, the resuspended proteins were reduced, alkylated and digested with trypsin (Promega, Southampton UK) according to the manufacturer’s protocol (Applied Biosystems, Foster City, CA). Biological duplicate protein samples were iTRAQ-labeled as follows (100 µg each per label): Day 0, 113; Day 1, 114; Day 3, 115; Day 6, 116; Day 0, 117; Day 1, 118; Day 3, 119; Day 6, 121 (Figure 1). The labeled samples were then pooled and dried in a vacuum concentrator prior to Strong Cation Exchange (SCX) fractionation. SCX Fractionation. Dried labeled peptides were resuspended in 200 µL of buffer A and fractioned using a PolySULFOETHYL A column (PolyLC, Columbia, MD) of 5 µm particle size of 200 mm length ×2.1 mm id, 200 Å pore size, on a BioLC HPLC unit (Dionex, Surrey, UK) with a constant flow rate of 0.2 mL/

research articles min and an injection volume of 200 µL. Buffer A consisted of 10 mM KH2PO4 and 25% acetonitrile, pH 3.0, and buffer B consisted of 10 mM KH2PO4, 25% acetonitrile, and 500 mM KCl, pH 3.0. The 60-min gradient consisted of 100% A for 5 min, 5-30% B for 40 min, 30-100% B for 5 min, 100% B for 5 min, and finally 100% A for 5 min. The chromatogram was monitored through a UV Detector UVD170U and Chromeleon Software, version 6.50 (Dionex/LC Packings, The Netherlands). The UV wavelengths were set at 280, 254, and 214 nm. Fractions were collected every minute using a Foxy Jr. Fraction Collector (Dionex) and later were pooled together according to variations in peak intensity. A total of 33 SCX fractions were pooled for subsequent nano-LC-MS/MS analysis. Pooled fractions were dried in a vacuum concentrator and stored at -20 °C prior to mass spectrometric analysis. Mass Spectrometric Analysis - ITRAQ. Each dried SCX iTRAQ-labeled peptide fraction was redissolved in 40 µL of 0.1% formic acid and 3% acetonitrile, and then 15 µL of sample was injected into the nano-LC-ESI-MS/MS system for analysis. Mass spectrometry was performed using a Q-Star XL Hybrid ESI Quadrupole time-of-flight tandem mass spectrometer, ESIQ-TOF-MS/MS (Applied Biosystems; MDS-Sciex), coupled to an online capillary liquid chromatography system (Famos, Switchos, and Ultimate from Dionex/LC Packings, Amsterdam, The Netherlands) as described elsewhere.19 The peptide mixture was separated on a PepMap C-18 RP capillary column (LC Packings), with a constant flow rate of 0.3 µL/min. The LC gradient started with 3% buffer B (0.1% formic acid in 97% acetonitrile) and 97% buffer A (0.1% formic acid in 3% acetonitrile) for 3 min, followed by 3% to 30% buffer B for 90 min, then 90% buffer B for 7 min, and finally 3% buffer B for 8 min. The mass spectrometer was set to perform data acquisition in the positive ion mode, with a selected mass range of 300-2000 m/z. Peptides with +2 to +4 charge states were selected for tandem mass spectrometry, and the time of summation of MS/ MS events was set to 3 s. The two most abundantly charged peptides above a 5 count threshold were selected for MS/MS and dynamically excluded for 60 s with (50 mmu mass tolerance. Protein Identification and Quantification. All database interrogation experiments were processed against a collapsed version of the NCBI nonredundant database containing only sequence entries from Homo sapiens (Feb 21, 2008; 197 944 unique Homo sapiens protein entries). LC-MS/MS analysis files for iTRAQ samples were searched via the Paragon algorithm using Protein Pilot v2.0 software (Applied Biosystems MDSSciex).20 Default instrument tolerances were specified for QSTAR mass spectrometers as provided within the Protein Pilot software. Miss cleavage tolerances were left as default (no limit). Methylthio (C; +46 Da) and iTRAQ 8-plex (N-term and K; +304 Da) were set as fixed modifications. LC-MS/MS analysis of ExacTag samples were searched via an in-house Phenyx server algorithm processing cluster at the ChELSI institute.21 ExacTag Thiol (C; +972 Da) was set as fixed modification and Oxidation (M; +16 Da) were selected as variable modification. Phenyx search mass tolerances were set to 1 Da MS and 800 ppm MS/ MS with at least 20% ion coverage. Miss cleavage tolerances were set at 1 with Trypsin as the primary enzyme. iTRAQ data filtering was performed to retain only protein identifications of g95% confidence interval and have multiple peptide identifications. ExacTag and Phenyx interrogation were filtered using a peptide z-score of 7.0 and p-val 0.05) were considered to be unaltered by differentiation over time. Gene Ontology. eGOnv2.0 software (http://www.genetools. microarray.ntnu.no/egon/index.php)23 was used to annotate proteins with Gene Ontology terms. eGOn used a two-sided Fisher’s exact test to determine whether any GO categories were significantly enriched in proteins that had been identified as being altered over the time course of differentiation (p < 0.05). To achieve this, a Master-Target test was performed: The master list contained all proteins identified (regardless of quantification status) by both iTRAQ and ExacTag. The target list contained only genes that were significantly altered over the time course of differentiation. Hierarchical Clustering. Proteins were clustered based on differences in expression patterns occurring over time using Cluster analysis software (http://rana.lbl.gov/EisenSoftware. htm).24 Cluster assembled a set of items (protein expression data) into a tree, and joined them by increasingly longer branches as their similarity decreased. The distance matrix between the expression data was calculated. Agglomerative hierarchical processing was then used, which consisted of repeated cycles where the two closest remaining items (those with the smallest distance) were joined by a node/branch of a tree, with the length of the branch set to the distance between the joined items. The two joined items were removed from list of items being processed and replaced by an item that represented the new branch. The distances between this new item and all other remaining items were computed, and the process was repeated until only one item remained. The output files were visualized with Java Treeview (http://jtreeview. sourceforge.net/docs/overview.html). Proteins represented within a cluster indicate similar distributions of expression values across experimental time points. Construction and Visualization of Interactome Networks. HiMAP (http://www.himap.org)25 was used to generate a human protein-protein interaction map. HiMAP is a dynamic browser for the human protein-protein interaction map, which utilizes literature-confirmed interactions from the Human Protein Reference Database, yeast-two-hybrid-defined interactions, and predicted interactions generated by a Bayesian Analysis. The accession numbers of significantly altered proteins were uploaded to HiMAP and probed against its database of 39 816 human protein-protein interactions. Thus, signifi3416

Journal of Proteome Research • Vol. 9, No. 7, 2010

Figure 2. Validation of nuclear protein enrichment method using Western blotting. Antibodies against nuclear Histone H3 and cytoplasmic Actin were used to semiquantitatively measure the nuclear protein enrichment in nuclear and cytoplasmic fractions. An increased abundance of Histone H3 was only detectable in the nuclear protein fraction, with almost no detectable Actin presence. No Histone H3 was detected in the cytoplasmic protein fraction. This validated the efficacy of nuclear extraction protocol for nuclear protein enrichment.

cantly altered proteins were explored for both known and predicted protein-protein interactions with each other. Analysis of Peptides. Unique peptides identified by iTRAQ and ExacTag were investigated for physiochemical properties. Peptide sequences were individually analyzed using the Innovagen peptide property calculator (http://www.innovagen.se/ custom-peptide-synthesis/peptide-property-calculator/peptideproperty-calculator.asp). The tool uses proprietary algorithms to calculate molecular weight, net charge, isoelectic point, and hydrophilicity of each peptide based on its amino acid composition. Experimental Design. To examine the proteomic profiles of NTERA-2 cells undergoing differentiation, NTERA-2 cells were incubated in the presence of 10-5 M all-trans-retinoic acid for either 0, 1, 3, or 6 days (Figure 1). At each of the four time points, a nuclear extraction was performed to isolate the nuclear proteins. Western blotting was performed to verify successful nuclear extraction and enrichment of nuclear proteins. To ascertain whether successful differentiation had occurred, the level of neuronal and pluripotency markers were measured in all groups using RT-PCR. Two isobaric-labeling proteomic technologies, ExacTag and iTRAQ were employed to study the proteomic profile of NTERA-2 undergoing RA-induced differentiation (Figure 1). This produced robust data regarding which nuclear proteins were influenced during differentiation. To verify the proteins identified by ExacTag and iTRAQ Western blotting was carried out as described above. For further insights into mechanisms of differentiation and comparison with proteomic analysis, RTPCR analyses were also performed.

Results Validation of Nuclear Protein Enrichment. To confirm that the nuclear extraction protocol had led to enrichment of nuclear protein within the samples, we performed Western blot

Nuclear Protein Dynamics of Differentiating NTERA-2 Cells

research articles

Figure 3. Identified proteins grouped based on Cellular Component using Gene Ontology tools. The majority of proteins belonged to nuclear, intracellular and chromosomal compartments with 32, 18, and 9 proteins respectively in each category; 32 proteins were nuclear, which further verified the efficacy of the nuclear extraction protocol for nuclear protein enrichment.

Figure 4. Identified proteins grouped based on Molecular Function using Gene Ontology tools. The majority of identified proteins were binding proteins (85 proteins). Of these, 33 were involved in protein binding, 33 were involved in nucleic acid binding and 17 were involved in nucleotide binding.

analysis using an antibody against Histone H3, which is not present in any of the cell compartments outside the nucleus. We also investigated to what extent residual cytoplasmic proteins may have been present in these fractions by using an antibody against cytoplasmic Actin. Western blot analysis confirmed an enrichment of Histone H3, and almost complete absence of cytoplasmic Actin in obtained nuclear protein fractions (Figure 2). The proteins identified in nuclear fractions by mass spectrometry were annotated using the gene ontology tool, eGOn v2.0 (http://www.genetools.microarray.ntnu.no/egon/index. php). Each protein was categorized based on cellular component (Figure 3) and molecular function (Figure 4). When grouped according to cellular component, the majority of identified proteins (32 proteins) were present in the nucleus (Figure 3). Also, when categorized based on molecular function (Figure 4), 33 proteins were grouped as being involved in nucleic acid binding and a further 17 proteins were implicated

in nucleotide binding. Relatively few proteins could be characterized as originating from potentially non-nuclear cellular compartments (5 cytoskeletal proteins, 1 cell-surface protein, 1 vesicle protein, and 1 golgi protein). Thus, our Western blots and GO analysis indicated that our nuclear protein fractionation strategy resulted in a significant enrichment of nuclear proteins. Differentiating NTERA-2 Cells Down-Regulated Pluripotency Markers and Up-Regulated Neuronal Markers. Monitoring the regulation of gene expression during NTERA-2 differentiation was performed to confirm that differentiation had occurred. RNA was isolated from undifferentiated NTERA-2 cells and their differentiated derivatives and subsequently prepared for gene expression analysis. Successful cell differentiation was assessed by the reduction of pluripotency-associated transcription factor expression, in parallel with up-regulation of neural-associated genes (Figure 5). Cells that were exposed to RA for 1 day exhibited downJournal of Proteome Research • Vol. 9, No. 7, 2010 3417

research articles

Pewsey et al.

Figure 5. Expression of neuronal and pluripotency markers during RA-induced differentiation. RT-PCR showed that cells exposed to RA exhibited down-regulation of the pluripotency associated transcription factor Oct4 (POU5F1) along with very low levels of Pax6 and Hash1 expression. The expression of NeuroD1, Pax6 and Hash1 positively correlated with the length of RA exposure. Eomesodermin expression did not follow the expression pattern exhibited by neural specific genes. Overall, the observed trend was a reduction of pluripotent transcription factor expression, in parallel with up-regulation of neural-associated genes.

regulation of the pluripotency-associated transcription factor Oct4 (POU5F1) whereas only very low levels of Pax6 and Hash1 expression were detected. Sox2 expression remained unchanged over the 6 day time course of differentiation. RAmediated reductions in the level of Oct4 mRNA were most prominent when cells were exposed to RA for 3 and 6 days. The expression of Eomesodermin, a T-box gene primarily associated with trophoblast and mesodermal development, did not follow the expression pattern exhibited by neural specific genes (Figure 5). It was modestly expressed by undifferentiated stem cells and its expression was specifically down-regulated in cell cultures that had been exposed to RA for the initial 24 h. The expression of NeuroD1, Pax6 and Hash1 positively correlated with the length of RA exposure, with the highest levels of neural specific gene expression found in cell cultures that had been continuously exposed to RA for the full 6 days. Proteomic Analysis of Nuclear Dynamics during Neuronal Differentiation. Extraction purified nuclear protein samples from Day 0, Day 1, Day 3 and Day 6 of neuronal differentiation were quantified by tandem mass spectrometrybased techniques using two independent isotope labeling technologies: ExacTag and iTRAQ.

ExacTag. Three biological replicates for each time point were pooled together to form one sample. Technical replication was assessed by duplicates of each time point, (pooled Day 0 was labeled with both isobaric tags 343 and 349, Day 1 was labeled with 355 and 400, Day 3 was labeled with 406 and 412 and Day 6 was labeled with 457 and 463) (Figure 1). From the tandem MS analysis, a total of 293 reliable protein identifications were obtained via Phenyx set cutoff (p-val