Proteome Analysis Reveals New Mechanisms of Bcl11b-loss Driven Apoptosis Narasimha Kumar Karanam,†,‡,§ Piotr Grabarczyk,†,| Elke Hammer,‡ Christian Scharf,‡,⊥ Simone Venz,‡,# Manuela Gesell-Salazar,‡ Winfried Barthlen,§ Grzegorz K. Przybylski,|,∇ Christian A. Schmidt,| and Uwe Vo ¨ lker*,‡ Interfakulta¨res Institut fu ¨ r Genetik und Funktionelle Genomforschung, Ernst-Moritz-Arndt-Universita¨t Greifswald, Germany, Klinik fu ¨ r Innere Medizin C, Universita¨tsklinikum der Ernst-Moritz-Arndt-Universita¨t Greifswald, Germany, Klinik und Poliklinik fu ¨ r Hals-, Nasen- und Ohrenkrankheiten, Kopf und Halschirurgie, Universita¨tsklinikum der Ernst-Moritz-Arndt-Universita¨t Greifswald, Germany, Institut fu ¨ r Medizinische Biochemie und Molekularbiologie, Ernst-Moritz-Arndt-Universita¨t Greifswald, Germany, Klinik und Poliklinik fu ¨ r Kinderchirurgie, Universita¨tsklinikum der Ernst-Moritz-Arndt-Universita¨t Greifswald, Germany, and Institute of Human Genetics, Polish Academy of Sciences, Poznan, Poland Received November 27, 2009
The Bcl11b protein was shown to be important for a variety of functions such as T cell differentiation, normal development of central nervous system, and DNA damage response. Malignant T cells undergo apoptotic cell death upon BCL11B down-regulation, however, the detailed mechanism of cell death is not fully understood yet. Here we employed two-dimensional difference in-gel electrophoresis (2DDIGE), mass spectrometry and cell biological experiments to investigate the role of Bcl11b in malignant T cell lines such as Jurkat and huT78. We provide evidence for the involvement of the mitochondrial apoptotic pathway and observed cleavage and fragments of known caspase targets such as myosin, spectrin, and vimentin. Our findings suggest an involvement of ERM proteins, which were up-regulated and phosphorylated upon Bcl11b down-regulation. Moreover, the levels of several proteins implicated in cell cycle entry, including DUT-N, CDK6, MCM4, MCM6, and MAT1 were elevated. Thus, the proteome data presented here confirm previous findings concerning the consequences of BCL11B knock-down and provide new insight into the mechanisms of cell death and cell cycle disturbances induced by Bcl11b depletion. Keywords: Bcl11b • ERM (Ezrin • Radixin • Moesin) • cell death • cell cycle disturbances • 2D-DIGE
1. Introduction The Bcl11b protein (B-cell chronic leukemia/lymphoma 11b), initially described as chicken ovalbumin upstream promoter transcription factor (COUP-TF)-interacting protein 2 (CTIP2)1 and radiation induced tumor suppressor gene 1 (RIT1),2 is involved in a variety of biological processes. They range from the regulation of T cell differentiation3 and * To whom correspondence should be addressed. Uwe Vo¨lker, ErnstMoritz-Arndt-Universita¨t, Interfakulta¨res Institut fu ¨ r Genetik and Funktionelle Genomforschung Friedrich-Ludwig-Jahn-Str. 15A, 17487 Greifswald, Phone: +49-3834-865870, Fax: +49-3834-8680012, E-mail: voelker@ uni-greifswald.de. † These authors contributed equally to this work. ‡ Interfakulta¨res Institut fu ¨ r Genetik und Funktionelle Genomforschung, Ernst-Moritz-Arndt-Universita¨t Greifswald. § Klinik und Poliklinik fu ¨r Kinderchirurgie, Universita¨tsklinikum der ErnstMoritz-Arndt-Universita¨t Greifswald. | Klinik fu ¨ r Innere Medizin C, Universita¨tsklinikum der Ernst-MoritzArndt-Universita¨t Greifswald. ⊥ Klinik und Poliklinik fu ¨r Hals-, Nasen- und Ohrenkrankheiten, Kopf und Halschirurgie, Universita¨tsklinikum der Ernst-Moritz-Arndt-Universita¨t Greifswald. # Institut fu ¨ r Medizinische Biochemie und Molekularbiologie, ErnstMoritz-Arndt-Universita¨t Greifswald. ∇ Polish Academy of Sciences. 10.1021/pr901096u
2010 American Chemical Society
normal development of central nervous system (CNS) during embryogenesis4,5 to DNA damage response6 and the maintenance of latent human immunodeficiency virus (HIV) infections.7 However, as recently shown, its role is not restricted to the central nervous and immune system. BCL11B is also essential for development of skin,8 where it regulates keratinocyte proliferation and the late differentiation phases determining the process of skin morphogenesis.9 Furthermore, normal tooth development was significantly impaired in BCL11B-deficient mice, which was accompanied by the decreased expression of ameloblast marker genes and transcription factors driving odontogenesis.10 Structurally, Bcl11b belongs to the C2H2-family of Krueppellike zinc finger proteins and thus is a member of the largest family of transcription factors in eukaryotes. Amino acids on the surface of an R helix contact a GC rich target sequence in the major groove and confer direct DNA-binding.11,12 Additionally, Bcl11b possesses transcriptional regulatory domains. Among the interaction partners identified so far are COUPTF,1 NuRD (nucleosome remodelling and histone deacetylation complex)13 and the ubiquitous transcription factor Sp1.14 It was shown that, as a consequence of recruiting both histone Journal of Proteome Research 2010, 9, 3799–3811 3799 Published on Web 05/31/2010
research articles
Karanam et al. 13,15
deacetylases (HDAC1 and HDAC2, respectively SIRT1) and a histone methyltransferase SUV39H1, Bcl11b induces heterochromatin formation and thus acts as a potent transcriptional repressor.7 In contrast, Bcl11b interaction with p300 coactivator on the upstream site 1 (US1) of the IL-2 promoter results in transcriptional activation and subsequently increased IL-2 expression after T cell activation.16 However, although interaction partners and the direct binding sequence have been described, besides IL-2 only few target genes of Bcl11b are known. One of them, the P57KIP2 gene, for example, a cyclindependent kinase inhibitor, is repressed by Bcl11b.17 Apart from P57 and IL-2 genes, the cancer Osaka thyroid oncogene (Cot) has been recently identified as a direct transcriptional target of Bcl11b. As a consequence, Cot expression was induced and its kinase activity enhanced the phosphorylation of IkappaB kinase, which led to increased translocation of NF-kappaB to the nucleus and activation of its target genes.18 Recently, Bcl11b was shown to bind the p21 promoter and in cooperation with SUV39H1 methyltransferase silenced the transcription of p21. This epigenetic silencing of p21 via the cooperative action of Bcl11b and SUV39H1 contributed indirectly to HIV latency in macrophages.19 The biological function of Bcl11b in vivo is still a matter of debate. In humans, overexpression of BCL11B mRNA has been linked to lymphoproliferative disorders like the T cell acute lymphoblastic leukemia, T-ALL20,21 and an acute form of adult T cell leukemia/lymphoma.22 Moreover, BCL11B up-regulation correlated with low differentiation status in head and neck squamous cell carcinoma. Interestingly, the Bcl11b protein colocalized with the cancer stem cell marker BMI-1.23 Conversely, in the mouse models of T cell leukemia, the BCL11B locus is frequently homozygously deleted or mutated and the expression of the gene is impaired.24 The loss of only one allele was identified as a factor predisposing to lymphoma development in p53 (() mice which suggests that BCL11B is a haploinsufficient suppressor for thymoma development in this genetic background and that deletion of one gene copy is advantageous for tumor growth.25 The mutated variants of murine BCL11B isolated from chemically induced T cell lymphomas in turn were able to enhance proliferation of a hematopoietic progenitor cell line in vitro.26 In conclusion, the contradictory results provided by animal models and human diseases intensify the dispute on the tumor suppressor or oncogenic properties of BCL11B. As we have shown previously, malignant T cell lines undergo apoptosis, when BCL11B expression is inhibited, which makes Bcl11b a potential therapeutic target in T cell derived tumors.27 Apoptosis is executed by simultaneous activation of receptor mediated and intrinsic apoptotic pathways and accompanied by a decrease in BCL-xL and an increase in Trail expression. This is in line with the apoptotic phenotype of BCL11B-/thymocytes that express lower levels of Bcl-xL and additionally Bcl-2.3 Moreover, a recent report by Kamimura et al.6 revealed that loss of Bcl11b leads to impaired response to DNA-damage and lack of cell cycle checkpoint activation accompanied by replication stress and cell death. These data stress the antiapoptotic function of Bcl11b but also document its role in maintaining genome stability which might contribute to the malignant transformation. The proteome data presented here not only confirm previous findings concerning the consequences of Bcl11B knock-down, 3800
Journal of Proteome Research • Vol. 9, No. 8, 2010
but most importantly, provide new insight into the mechanisms of cell death and cell cycle disturbances induced by Bcl11b depletion.
2. Materials and Methods 2.1. Cell Culture. Jurkat human T cell leukemia (DSMZ, Braunschweig, Germany) and huT78 human T cell lymphoma cell lines (ATCC, Manassas, VA) were grown in Rosewell Park Memorial Institute medium (RPMI) 1640 medium (Invitrogen, Paisley, U.K.) supplemented with 1 mM sodium pyruvate, 1× nonessential amino acids (Invitrogen), 10% fetal calf serum (FCS) (Invitrogen) and 5 mg/mL Plasmocin (Amaxa, Cologne, Germany) and maintained in a humidified incubator at 37 °C and 5% CO2. All experiments were performed using cells in exponential phase of growth. The retrovirus producing cell line GP2-293 (Clontech Laboratories, Mountain View, CA) was maintained in Dulbecco’s modified Eagle’s medium (Invitrogen, CA, USA) supplemented as above. 2.2. siRNA and Cell Transfection. BCL11B-specific (BCL11B674) stealth siRNAs and the corresponding nonsilencing control (sc) RNA duplexes were purchased from Invitrogen. Five micrograms of each siRNA were used for each transfection. Cells were transfected with the Nucleofection Device II (Amaxa), using the nucleofection Kit V (Amaxa) according to the manufacturer’s instructions. The BCL-XL coding sequence was amplified and ligated into XhoI and HindIII sites of the pIRES2EGFP vector (BD Clontech, Heidelberg, Germany) using T4DNA Ligase (Promega, Madison, WI). For stable transfection, Jurkat and huT78 cells were transfected with the BCL-XL containing vector followed by selection in the presence of 1 mg/mL G418 (Geneticin, Invitrogen). A recombinant retroviral vector was used to overexpress BCL11B. The infectious media were prepared by cotransfection of nonconfluent GP2-293 cells with 10 µg of purified pVSV-G and empty pMIGR plasmid (mock) or pMIGR encoding BCL11B. The transfection procedure was performed using CalPhos Mammalian Transfection Kit (Clontech Laboratories, Mountain View, CA) according to the manufacturer’s instructions. The retrovirus-containing media were collected 48 h after transfection and used immediately or stored at 4 °C and used within 7 days. To transduce the target cells, the retroviral supernatants were supplemented with 8 µg/ mL Polybrene (Sigma Chemicals, Germany) and added to exponentially growing Jurkat cells for 8 h. The procedure was repeated three times. The efficiency of gene transfer was estimated by FACS detection of EGFP reporter gene. 2.3. Determination of Apoptosis. Annexin-V-binding assays were performed at 24, 48, and 72 h after siRNA nucleofection using the BD Clontech ApoAlert Annexin-V-FITC or AnnexinV-APC (BD Pharmingen, Erembodegem, Belgium) kits according to the manufacturer’s instructions. The percent of apoptotic cells was quantified by FACS analysis as previously described.27 2.4. Cytofluorometric Analysis of Mitochondrial Transmembrane Potential (∆Ψm). The accumulation of the cationic lipophilic fluorochrome 3,3′-dihexyloxacarbocyanine iodide [DiOC6(3)] in the mitochondrial matrix is directly proportional to mitochondrial potential (∆Ψm). ∆Ψm was determined at 3 different time points at 24, 48 and 72 h after cell transfection. Cells were incubated with 50 nM DiOC6 (Molecular Probes; distributed by Mobitec, Go¨ttingen, Germany) at 37 °C for 30 min.28 After washing, at least 10 000 cells were analyzed using a Becton Dickinson FACScalibur and CellQuest and WinMDI2.8
New Mechanisms of Bcl11b-loss Driven Apoptosis
research articles
software (Becton Dickinson, Heidelberg, Germany). Data were gated to exclude debris.
Molecular imager FX (Biorad, Munich, Germany) at the excitation/emission wavelength of 532/555 nm.
2.5. Preparation of Protein Extracts and Two-Dimensional Difference in-Gel Electrophoresis (2D-DIGE). The cell pellet was resuspended in 2D rehydration buffer (8 M Urea, 2 M Thiourea, 2% CHAPS) and lysed by sonication 3 times for 5 s each with 9 cycles at 50% energy (Bandelin Electronics, Berlin, Germany). After removal of cell debris by centrifugation (20 000 × g for 20 min at 4 °C), concentration of the crude protein extracts was determined according to Bradford.29 Prior to 2D-DIGE, protein extracts were labeled with Cy-dyes as previously described.30 The Cy2 dye was used for labeling of a pooled standard of all samples included in the experiment. The 2D-DIGE with 24 cm IPG strips of the pH range 4-7 and the subsequent scanning were done according to published protocols.30 Each crude protein extract was analyzed on four 2-DE gels (four technical replicates) using two Cy3 and two Cy5 labeled samples thus enabling correction of dye-specific effects in the subsequent statistical analysis. 2.6. Data Analysis of 2D-DIGE Experiments. After scanning of gels with a Typhoon scanner (GE Healthcare, Munich, Germany), analysis of the 2-DE images was performed with version 3.4 of the Delta2D software package (Decodon GmbH, Greifswald, Germany). All gel images were matched with the Delta2D software and fused to generate a synthetic fusion gel using the union option. Subsequently, spot detection was performed for the fusion gel and the spot map of the fusion gel was then propagated to each of the three individual 2-DE images (Cy2, Cy3and Cy5) of each of the gels of the experiment thus allowing uniform spot quantification on all gels included in the study. After background subtraction, spot volumes were calculated with the Delta2D software and normalized both to the corresponding image (relative expression of each spot compared to the total spot volume of the image) and the pooled Cy2 standard of each individual spot (ratio of the particular sample spot volume to the Cy2 spot volume of the internal standard). The resulting values, which were normalized for image and pooled standard (percentage volume as defined by the Delta2D software), were exported to the GeneSpring software version 7.3.1 (Agilent Technologies, Waldbronn, Germany). Protein levels were considered to be changed when the following two criteria were fulfilled: (i) changes in the level of the protein had to be statistically significant, as defined in a statistical group comparison of the values of the selected conditions with a parametric test (Welch t test, p value cutoff 0.05), as defined in the GeneSpring software package; and (ii) the change in level had to exceed a factor of 1.5.
After Pro-Q Diamond phosphoprotein staining the same gels were washed with distilled water for 1 h with 3 changes and fixed in 40% ethanol and 10% acetic acid for 1 h, which was followed again by washing with distilled water 3 times for 10 min each. Then, the gels were stained with SyproRuby (Biorad) overnight. Subsequently, gels were rinsed in 10% methanol and 7% acetic acid for 1 h with 3 changes. Finally, the gels were washed with water prior to scanning using a Molecular imager FX at the excitation/emission wavelength of 450/610 nm. The images of the Pro-Q Diamond and SyproRuby stains, which reflect protein phosphorylation and total protein level, respectively, were matched using version 3.4 of the Delta2D software (Decodon GmbH) as described above. The phosphorylation signal was considered specific and significant when the signal in the Pro-Q Diamond stained image was stronger than that of the SyproRuby stained image. 2.8. Identification of Proteins by MALDI-MS/MS. Differentially expressed spots were manually excised from preparative 2D gels, stained using a colloidal CBB-staining procedure,31 digested with trypsin and the resulting peptides were spotted onto a MALDI-target using an Ettan Spot Handling Workstation (GE Healthcare) according to a standard protocol described previously.32 The matrix-assisted laser desorption/ionization time-of-flight mass spectrometry (MALDI-TOF MS/MS) of spotted peptide solutions was carried out using a 4800 Proteomics Analyzer (Applied Biosystems, Foster City, CA). MS analysis was performed with the following settings: 10 subspectra with 100 shots per subspectrum were accumulated using a random search pattern and a mass range of m/z 804 up to 4000. An internal calibration was automatically performed as a one or two point calibration for self-digested trypsin fragments at m/z 1045.5 and m/z 2211.1. MS/MS analysis was performed for the 5 strongest peaks of the MS spectrum. For one main spectrum, 20 subspectra with 125 shots per subspectrum were accumulated using a random search pattern. The internal calibration was automatically performed as one-point calibration if the monoisotopic arginine (M + H)+ m/z at 175.119 or lysine (M + H)+ m/z at 147.107 reached a signalto-noise (S/N) ratio of at least 20. The peak lists were created using the GPS Explorer software (Applied Biosystems) with the following settings: mass range from 60 Da to a mass that was 20 Da lower than the precursor mass; peak density of 10 peaks per 200 Da; minimal area of 100 and maximal 100 peaks per precursor; minimal S/N ratio of 7. Database searches employed a Homo sapiens specific database (Swiss-Prot Version 51.5, IPI human Version 3.23 respectively) using the Mascot search engine (Ver. 2.0; Matrix Science Ltd., London, U.K.). The identification of a protein spot was considered significant if the Mowse score exceeded a value of 60, which corresponds to a p value of 0.05. 2.9. Identification of Proteins by ESI-MS/MS. Candidate 2D gel spots that could not be identified using the MALDIMS/MS approach were subsequently subject to analysis by ESI-MS/MS. After digestion with sequence grade trypsin (Promega, Madison, WI) overnight, peptides were extracted using 50% acetonitrile containing 1% acetic acid in a water bath sonicator. The MS analysis of the peptide mixture was carried out using a LTQ-Orbitrap (Thermo Electron, Bremen, Germany) coupled to nano Acquity UPLC system (Waters Inc., Manchester, U.K.). Chromatographic separation of peptide mixture was done using a 100 µm × 100 mm Nano acquity UPLC column
2.7. Assessment of Protein Phosphorylation by Staining with Pro-Q Diamond. Crude protein extracts (150 µg each) of the scrambled control and siRNA samples were subjected to isoelectric focusing on immobilized pH gradient (IPG) strips (24 cm, pH-range: 4 to 7; GE Healthcare) followed by separation according to molecular mass in the second dimension as described above. After 2-DE gels were fixed in 50% [v/v] ethanol-2% [v/v] acetic acid solution for 1 h and washed with distilled water for 4 times, 15 min each. Then, gels were incubated in Pro-Q Diamond phosphoprotein staining solution (Molecular Probes, Eugene, OR) for 2 h in the dark. After staining gels were destained using 50 mM sodium acetate pH4 containing 20% propylene glycol for 3 times, 30 min each. Prior to the scanning gels were washed with distilled water for two times, 10 min each. Then, the gels were scanned using a
Journal of Proteome Research • Vol. 9, No. 8, 2010 3801
research articles at a constant temperature of 35 °C. Following a 3 min washing using solution A (2% acetonitrile, 0.1% acetic acid) peptides were eluted using a 35 min linear gradient to 60% solution B (100% acetonitrile, 0.1% acetic acid) with a flow rate of 0.9 µL/ min. Data acquisition and reduction was performed with Xcalibur version 2.0 (Thermo Electron, San Jose, CA). Protein identification search was done on a Sequest cluster (Thermo Electron, San Jose, CA) using Swiss-Prot human database version 51.5. 2.10. Western-Blot Analysis. Aliquots of protein extracts (30 or 50 µg) were separated by SDS-PAGE and proteins were blotted onto a pre-equilibrated Immobilon PVDF membrane (Millipore, Schwalbach, Germany) using a semidry western transfer apparatus (GE Healthcare). Equal protein loads were confirmed by immunodetection of GAPDH. Unspecific binding to the membranes was blocked by 5% nonfat dry milk powder (Carl Roth GmbH, Karlsruhe, Germany) in TBST solution (20 mM Tris, 138 mM NaCl, pH7.6 containing 0.1% Tween20) (Sigma, Munich, Germany). After blocking, membranes were washed with TBST and incubated with primary antibodies i.e. BCL11b (1:5000; Bethyl, Montgomery, TX), phospho ERM (1: 1000), total ERM (1:1000; Cell signaling, Danvwers, MA, USA), DUT-N (1:1000; Abcam, Cambridge, U.K.), PDCD5 (1:1000; Abcam, Cambridge, U.K.), MNAT1 (1:1000; Santacruz Biotechnology, Santa Cruz, CA), UCK2 (1:2000; Abcam, Cambridge, U.K.), and GAPDH (1:1000; Santacruz Biotechnology, Santa Cruz, CA) overnight in TBST solution containing 2% BSA. After washing, membranes were incubated with the corresponding alkaline phosphatase conjugated secondary anti rabbit or mouse antibodies (Dianova GmbH, Hamburg, Germany) and antibody specific binding was visualized by NBT/BCIP (Carl Roth GmbH,) or super signal west femtomaximum sensitivity (Pierce, Rockford, IL) substrate solutions using a Lumi-Imager instrument (Roche GmbH, Mannheim, Germany). Signal intensities were quantified using the Imagequant software version 5.2 (GE Healthcare) and protein specific intensities were corrected for the corresponding GAPDH signal. Western blotting experiments were performed for three biologically independent samples sets. For statistical evaluation of the analysis GAPDH-corrected intensity values were set to one and the ratios of the intensities were calculated by dividing the values of the BCL11B knock-down samples by those of the controls. Thereafter, averages and standard deviations were calculated and displayed as SEM ( SD. Statistical significance of the differences was assessed with an unpaired t test and marked with the asterisk (*) designating p values 1.5, p-value
