The International Journal of Biochemistry & Cell Biology 34 (2002) 1059–1070
Voltage-dependent anion-selective channel (VDAC) interacts with the dynein light chain Tctex1 and the heat-shock protein PBP74 Christian Schwarzer, Shitsu Barnikol-Watanabe, Friedrich P. Thinnes∗ , Norbert Hilschmann Max-Planck-Institute for Experimental Medicine, Department of Immunochemistry, Hermann-Rein Street 3, 37075 Göttingen, Germany Received 1 October 2001; received in revised form 6 February 2002; accepted 8 February 2002
Abstract The voltage-dependent anion-selective channel 1 (VDAC1), i.e. eukaryotic porin, functions as a channel in membranous structures as described for the outer mitochondrial membrane, the cell membrane, endosomes, caveolae, the sarcoplasmatic reticulum, synaptosomes, and post-synaptic density fraction. The identification of VDAC1 interacting proteins may be a promising approach for better understanding the biological context and function of the channel protein. In this study human VDAC1 was used as a bait protein in a two-hybrid screening, which is based on the Sos recruitment system (SRS). hVDAC1 interacts with the dynein light chain Tctex-1 and the heat-shock protein peptide-binding protein 74 (PBP74)/mitochondrial heat-shock protein 70 (mtHSP70)/glucose-regulated protein 75 (GRP75)/mortalin in vivo. Both interactions were confirmed by overlay-assays using recombinant partner proteins and purified hVDAC1. Indirect immunofluorescence on HeLa cells indicates a co-localisation of hVDAC1 with the dynein light chain and the PBP74. In addition, HeLa cells were transfected transiently with enhanced green fluorescent protein (EGFP)–hVDAC1 fusion proteins, which also clearly co-localise with both proteins. The functional relevance of the identified protein interactions was analysed in planar lipid bilayer (PLB) experiments. In these experiments both recombinant binding partners altered the electrophysiological properties of hVDAC1. While rTctex-1 increases the voltage-dependence of hVDAC1 slightly, the rPBP74 drastically minimises the voltage-dependence, indicating a modulation of channel properties in each case. Since the identified proteins are known to be involved in the transport or processing of proteins, the results of this study represent additional evidence of membrane-associated trafficking of the voltage-dependent anion-selective channel 1. © 2002 Elsevier Science Ltd. All rights reserved. Keywords: VDAC; Porin; Tctex-1; PBP74; Two-hybrid system; Trafficking; Chloride channels
Abbreviations: aa, amino acid; EGFP, enhanced green fluorescent protein; FITC, fluorescein isothiocyanate; GABAA , ␥-aminobutyric acid; GRP75, glucose-regulated protein 75; GST, glutathione S-transferase; HRP, horseradish peroxidase; hVDAC1, human voltage-dependent anion-selective channel 1; LMW, low molecular weight marker; mtHSP70, mitochondrial heat-shock protein 70; ORF, open reading frame; PBP74, peptide-binding protein 74; PLB, planar lipid bilayer; PVDF, polyvinylidene difluoride; RVD, regulatory volume decrease; SDS-PAGE, sodium dodecylsulfate polyacrylamide gel electrophoresis; SRS, Sos recruitment system; TBS, Tris buffered saline; TRITC, tetramethylrhodamine isothiocyanate ∗ Corresponding author. Tel.: +49-551-3899-373; fax:+49-551-3899-500. E-mail address:
[email protected] (F.P. Thinnes). 1357-2725/02/$ – see front matter © 2002 Elsevier Science Ltd. All rights reserved. PII: S 1 3 5 7 - 2 7 2 5 ( 0 2 ) 0 0 0 2 6 - 2
1060
C. Schwarzer et al. / The International Journal of Biochemistry & Cell Biology 34 (2002) 1059–1070
1. Introduction The voltage-dependent anion-selective channel 1 (VDAC1), i.e. eukaryotic porin, is an abundant protein of the outer mitochondrial membrane. The channel forms a major pathway for the movement of adenine nucleotides, exhibits a high-conductance anion-selective open state, and cation-selective substates at applied voltages > ±20 mV in PLB experiments [1]. Apart from its role in energy metabolism and mitochondrial homeostasis, recent studies suggest that VDAC1 is involved in apoptotic cell death in this compartment. The release of cytochrome c through VDAC1 is modulated by pro- and anti-apoptotic members of the Bcl2 family (Bax and BclXL ) [2,3]. In addition, VDAC1 was found to be not only expressed in the mitochondria, but also in the plasma membrane [4], the sarcoplasmatic reticulum [5,6], and in endosomes [7]. The extramitochondrial expression of VDAC1, identified first immunologically, has been discussed controversially because of a putative cross-reactivity of the applied antibodies [8]. Recent studies performed in different laboratories confirm the plasmalemmal expression of VDAC1 using alternative methods: The channel was identified in caveolae of neoplastic hematopoetic CEM cells using extracellularly applied NH-SS-biotin [9], Myc-tagged mouse VDAC molecules were expressed at the cell membrane of transfected COS7 and HeLa cells [10]. Finally, hVDAC1-EGFP fusion proteins and FLAG-tagged hVDAC1 could be expressed heterologously at the cell membrane of Xenopus oocytes [11]. The plasmalemmal VDAC forms a large conductance anion channel with electrophysiological characteristics similar to VDAC isolated from mitochondria, and studied in PLB experiments [10,12,13]. The effects of anti-VDAC1 antibodies on the gadolinium induced swelling behaviour or the RVD of B-lymphocytes and HeLa cells suggest the involvement of the plasmalemmal VDAC1 in cellular volume regulation [14,15]. To function as a channel protein in different compartments with varying requisitions, a specific regulation of VDAC1 is essential. Protein–protein interaction is one way to modulate electrophysiological channel properties. In addition, protein-sorting within the cell is controlled by proteins.
In the present study, hVDAC1 was used as bait protein in a yeast two-hybrid screening in order to identify new interacting proteins. The identification of protein–protein interaction between hVDAC1 and the dynein light chain Tctex1 on the one hand, and the heat-shock protein PBP74/mtHSP70/GRP75/mortalin (here termed as PBP74) on the other hand, presents new insights on the biological context of hVDAC1. 2. Materials and methods The applied restriction enzymes and ligase were supplied by Life Technologies GmbH, Karlsruhe, Germany. Vent polymerase (New England Biolabs GmbH, Frankfurt, Germany) was used for all PCR amplifications. All cloning steps were controlled by means of DNA-sequencing using ABI PrismTM 377 DNA Sequencer (PE Biosystems Deutschland GmbH, Weiterstadt, Germany). 2.1. cDNA library construction mRNA was prepared from the lymphoid B cell line H2LCL using a Messenger RNA Isolation Kit (Stratagene, La Jolla, USA). The purified mRNA from 5×107 cells was used to synthesise hemimethylated cDNA (CytoTrapTM XR Library Construction Kit, Stratagene, La Jolla, USA). The cDNA was ligated into the EcoRI- and XhoI-restriction sites of the vector pMyrXR (Stratagene, La Jolla, USA). E. coli XL10 Gold Kan cells (Stratagene, La Jolla, USA) were transformed with the corresponding plasmids and amplified once. Plasmids were isolated and used in two-hybrid screening. 2.2. CytoTrapTM two-hybrid system A two-hybrid system based on the SRS [16] was used to screen a human lymphoid cDNA library for hVDAC1 interacting proteins according to the manufacturer’s instructions (Stratagene, La Jolla, CA, USA). The full length ORF of hVDAC1 was amplified by PCR from a human pituitary cDNA (a kind gift from M. Forte, Vollum Institute, Portland, Oregon, USA). Primers were composed of the 5 - and 3 -end of the ORF from hVDAC1 and contained BamHI and SalI restriction sites, respectively. The amplified PCR
C. Schwarzer et al. / The International Journal of Biochemistry & Cell Biology 34 (2002) 1059–1070
product was cloned in frame into the vector pSos. SRS screening was conducted by co-transformation of cdc25H yeast cells with pSos-hVDAC1 and pMyr- library plasmids. Co-transformation of cdc25H cells with pSos-hVDAC1/pMyr, pSos-hVDAC1/pMyrLamin C, or pSos/pMyr functioned as negative controls, whereas pSos-MafB/pMyr-MafB served as a positive control. After co-transformation and plating of cdc25H cells on the appropriate media, plasmid DNA was isolated from colonies that showed galactose-dependent growth at the restrictive temperature (37 ◦ C). Protein interactions of putatively positive colonies were confirmed by co-transformation of cdc25H cells, using each remaining pMyr-plasmids with either pSos-hVDAC1 or pSos. Only those clones in which interaction of Sos-hVDAC1 (but not Sos alone) took place were defined as positive. The pMyr plasmids of these putatively positive clones were isolated and the derived cDNA sequence was used to search the GenBank database with the BLAST program. 2.3. Recombinant protein expression (GST-fusion proteins) To confirm the in vivo interaction of hVDAC1 with Tctex-1 and PBP74/mtHSP70/GRP75/mortalin, recombinant proteins were synthesised. Therefore, the ORF of Tctex-1 was amplified from a H2LCL cDNA library by PCR using the forward primer 5 -CGCGTGGATCCATGGAAGACTACCAGGCTGCG-3 and the reverse primer 5 -AGCTTGTCGACTCAAATAGACAGTCCGAAGGCAC-3 . The ORF encoding for human PBP74/mtHSP70/GRP75/mortalin was amplified by PCR using forward primer 5 -GCGTGGATCCATGATAAGTGCCAGCCGAGCTGCA-3 and reverse primer 5 -TCGAGTCGACTTACTGTTTTTCCTCCTTTTGATC-3 . For PCR amplification of the partial cDNA sequence of human PBP74/mtHSP70/ GRP75/mortalin (which was identified through the SRS), the forward primer consisted of 5 -GCGTGGATCCATGATTGAAACTCTAGGAGGTGTC-3 . The PCR products were cloned in frame into the BamHI and the SalI restriction sites of pGEX-KG [17]. E. coli BL21(DE3)pLys (Novagen Inc., Madison, WI, USA) were transformed with the corresponding plasmids and each GST-fusion protein was isolated using affinity chromatography on GlutathioneSepharose® 4B according to the manufacturers
1061
instructions (Pharmacia LKB, Freiburg, Germany). The identity and quality of the recombinant proteins was verified using SDS-PAGE and western blot analysis. 2.4. Overlay-assay For verification of protein–protein interactions, hVDAC1 and each recombinant protein were separated by SDS-PAGE and transferred electrophoretically to a PVDF membrane. SDS-PAGE standards were purchased from Amersham Biosciences, Freiburg, Germany and Bio-Rad Laboratories GmbH, München, Germany. After the blocking of nonspecific sites with 0.5% BSA, 0.25% Tween 20 in TBS (25 mM Tris–HCl pH 7.4, 150 mM NaCl) for 2 h at room temperature, and washing five times with 0.5% Tween 20 in TBS, the membrane was incubated for 1 h at room temperature or overnight at 4 ◦ C with a probe of the recombinant protein (20 g/ml in TBS). The membrane was then washed five times in TBS and incubated with antibodies against the probe protein, followed by HRP-conjugated secondary antibodies. The blot was developed using 4 ml TBS, 2 ml 0.3% 4-chloro-1-naphthol (w/v), and 5 l H2 O2 (30%). Overlay-assays were performed twice. 2.5. Transfection and immunofluorescence Analysis of the subcellular distribution of hVDAC1 and the interacting proteins was done using native and EGFP–hVDAC1 transfected HeLa cells. To construct recombinant plasmids for the expression of hVDAC1 and EGFP fusion proteins, the ORF of hVDAC1 was amplified by PCR using the human pituitary cDNA described above. The forward primer encoded the 5 -end of the ORF and a XhoI restriction site. The reverse primer was composed of a PstI restriction site at the 3 -end of the ORF without a stop codon for the synthesis of hVDAC1-EGFP fusion proteins, and with a stop codon for EGFP–hVDAC1 fusion proteins. The amplified PCR products were cloned in frame into pEGFP-C3 and pEGFP-N1 (Clonetech Laboratories GmBH, Heidelberg, Germany). Transfection of HeLa cells with EGFP constructs was performed with the Superfect Transfection Reagent according to the manufacturer’s instructions (QIAGEN, Hilden, Germany).
1062
C. Schwarzer et al. / The International Journal of Biochemistry & Cell Biology 34 (2002) 1059–1070
For indirect immunofluorescence, native or transiently transfected HeLa cells were cultured on coverslips for 24–48 h. Cells were washed in TBS and fixed with 4% paraformaldehyde followed by incubation with 0.3% Triton X100 in PBS, or alternatively air-dried and fixed with 80% acetone in PBS. Cells were washed and blocked by 1% BSA in PBS and incubated for 60 min with a 1:50 dilution of monoclonal anti-hVDAC1 (Calbiochem-Novabiochem, Bad Soden/Ts., Germany) or anti-PBP74/mtHSP70/ GRP75/mortalin antibodies (Affinity Bioreagents Inc., Golden, USA). Polyclonal antisera anti-Tctex-1 (a kind gift of Stephan King, Department of Biochemistry, University of Connecticut, USA) and antisera against the acetylated N-terminus of hVDAC1 [4] were diluted 1:10. Secondary FITC-conjugated sheep anti-mouse, TRITC-conjugated sheep anti-mouse, FITC-conjugated sheep anti-rabbit, and TRITC-conjugated sheep anti rabbit antibodies (Sigma, Taufkirchen, Germany) were applied for 45 min at a dilution ratio of 1:50. Nuclear staining was performed using the supervital dye H33342 (Hoechst, Frankfurt, Germany). Coverslips were transferred upside down on object slides and sealed with rubber cement to prevent drying. Fluorescent labelling was analysed using the Axioskop fluorescent microscopy system (Zeiss, Göttingen, Germany). Experiments were performed three times and representative images were documented. 2.6. PLB experiments PLB experiments were performed as described elsewhere [18]. Channel-active hVDAC1 was purified from H2LCL cells by two steps of ion-exchange chromatography [4]. The interaction of hVDAC1 with the in vivo identified partner proteins was analysed after pre-incubation 50 ng of hVDAC1 (cholesterol saturated) with 1 g of the recombinant proteins for 30 min at room temperature in 1× PBS. In addition, the effect of 1 g of the recombinant proteins on 50 ng membrane-inserted (cholesterol saturated) hVDAC1 was analysed. Probes were applied symmetrically to both chambers. The membrane consisted of l-␣-diphytanoyl phosphatidylcholine (Avanti Polar Lipids Inc., Alabaster, AL, USA). Channel insertion was conducted at a membrane potential of +10 mV
until the membrane was saturated with hVDAC1. All measurements were done in ringer solution (145 mM NaCl, 5 mM KCl, 2 mM CaCl2 , 1 mM MgCl2 , 10 mM glucose, 10 mM Hepes, adjusted to pH 7.4 with NaOH). The relative conductance was calculated by dividing the conductance G at an applied voltage by the conductance G0 at + 10 mV. 3. Results VDAC1 is a channel protein expressed in membranes facing the cytoplasm. Therefore, identification of interacting proteins was performed by a yeast two-hybrid system based on the SRS [16]. This system enabled the detection of hVDAC1-interacting proteins anchored at the cytoplasmatic site of the plasma membrane. Protein-binding was analysed by means of protein chemistry and electrophysiology. 3.1. Identification of interacting proteins A cDNA library was constructed from the lymphoid B cell line H2LCL. Yeast two-hybrid screening was performed with hVDAC1 as a bait protein using the CytoTrapTM two-hybrid system. For this purpose, the temperature-sensitive yeast strain cdc25H was simultaneously transformed with pMyr-cDNA and pSos-hVDAC1 plasmids. Transformed yeast cdc25H cells with specific galactose-induced growth at the permissive temperature (37 ◦ C) were selected. pMyr plasmids of these clones were isolated and used together with either pSos-hVDAC1 or pSos (without an insert) for simultaneous retransformation of yeast cdc25H cells. Cells transformed with pSos-hVDAC1 (but not with pSos alone), showing only galactose-induced growth at 37 ◦ C, were defined as putatively positive clones. Fig. 1a–c shows the galactose-dependent growth of positive colonies transformed with pSos alone (four left plaques per line) and with pSos-hVDAC1 (four right plaques per line). For cells transformed simultaneously with pSos-hVDAC1 and specific pMyr-plasmids, colony growth was visible on glucose medium at 25 ◦ C (Fig. 1a), on galactose medium at 37 ◦ C (Fig. 1c), but not on glucose medium at 37 ◦ C (Fig. 1b). Control transformations (Fig. 1a–c four left plaques and Fig. 1d–f) showed the expected cell growth on glucose medium at 25 ◦ C
C. Schwarzer et al. / The International Journal of Biochemistry & Cell Biology 34 (2002) 1059–1070
1063
Fig. 1. Interaction-dependent growth of transformed yeast cells using the CytoTrapTM two-hybrid system. pMyr and pSos plasmid combinations were used for the simultaneous transformation of cdc25H yeast cells as listed beside the figures. Transformants were plated on glucose medium at 25 ◦ C (a) and (d), and replica were made on glucose medium at 37 ◦ C (b) and (e) and on galactose medium at 37 ◦ C (c) and (f). Colony growth of transformed yeast cells, in which interaction of hVDAC1 with specific proteins were observed, are seen in (a), (b), and (c). The control transformation reactions are presented in (d), (e), and (f).
and no growth on glucose or galactose media at 37 ◦ C (except for pMyr-MafB + pSos-MafB as the positive control). By screening 2 × 106 clones, the cDNAs of 13 positive clones were analysed by means of DNA-sequencing, and identified using the GenBank BLAST program. The cDNA of these pMyr plasmids encoded for the dynein light chain Tctex-1 (GenBank Accession #: D50663, aa 1-113) and a C-terminal portion of the (PBP74/mtHSP70/GRP75/mortalin) (GenBank Accession #: L11066, aa 448–679). 3.2. Confirmation of protein–protein interactions Verification of protein-binding to isolated hVDAC1 was performed with overlay-assays using recombinant Tctex-1 and PBP74. For this purpose, GST fusion proteins containing Tctex-1, PBP74, or part of PBP74 (aa 448–679, termed PBP74) were expressed
in E. coli, purified on glutathione sepharose, and eluted with thrombin. Isolated hVDAC1 and the respective recombinant proteins were loaded onto SDS-PAGEs (Fig. 2a upper panel for Tctex-1, Fig. 2b upper panel for PBP74, Fig. 2c upper panel for PBP74) and transferred onto PVDF membranes. Here, binding to hVDAC1 by rTctex-1, rPBP74, or rPBP74 was visualised using specific antibodies against Tctex-1 or PBP74, respectively, followed by enzymatic colour reaction (Fig. 2a lower left panel for Tctex-1, Fig. 2b lower left panel for PBP74, Fig. 2c lower left panel for PBP74). In contrast, no cross-reactivity between hVDAC1 and the applied antibodies was observed on control blots incubated only in buffer (Fig. 2a lower right panel for Tctex-1, Fig. 2b lower right panel for PBP74, Fig. 2c lower right panel for PBP74). The 30 kDa mass-marker protein was stained in Fig. 2b, lanes 1 and 4, lower panel. Because this protein is
1064
C. Schwarzer et al. / The International Journal of Biochemistry & Cell Biology 34 (2002) 1059–1070
Fig. 2. Overlay assay of hVDAC1 with recombinant partner proteins. Isolated hVDAC1 and each recombinant protein were separated on SDS-PAGE (coomassie stained control gel, upper panels) and transferred to a blot membrane (lower panels). The blots were incubated with the appropriate recombinant protein (left panels), or with TBS alone (right panels), followed by specific antibodies to the recombinant protein and HRP-conjugated secondary antibodies. (a) hVDAC1 and rTctex1, lane 1 and 4: LMW, lane 2 and 5: hVDAC1, lane 3 and 6: rTctex-1. (b) hVDAC1 and PBP74, lane 1 and 4: LMW, lane 2 and 5: hVDAC1, lane 3 and 6: PBP74. (c) hVDAC1 and PBP74, lane 1 and 4 upper panel: LMW, lane 1 and 4 lower panel: SDS-PAGE standard, lane 2 and 5: hVDAC1, lane 3 and 6: rPBP74.
also seen on the control blot, staining is based on the applied antibodies and not on unspecific binding of PBP74 to the mass-marker protein. Based on these experiments, the in vivo interaction of hVDAC1 with Tctex-1, PBP74, and the
entire PBP74 could be confirmed through direct protein–protein binding in vitro. Binding of the identified recombinant protein to hVDAC1 involved participation of residues within an approximately 3–4 kDa large C-terminal segment of hVDAC1. This emerges
C. Schwarzer et al. / The International Journal of Biochemistry & Cell Biology 34 (2002) 1059–1070
1065
from the failed direct binding of the interacting proteins to hVDAC1, which is missing this C-terminal segment. This protein band which separates directly below the hVDAC1 band (as seen in the SDS-PAGEs in Fig. 2a–c) represents a degradation product. Antibodies against the hVDAC1 N-terminus recognised this protein band in western blot experiments, indicating that a C-terminal segment is missing (data not shown). Recombinant partner proteins did not bind to that hVDAC1 fragment (Fig. 2a–c).
acetone-fixed HeLa cells (Fig. 3e–g and Fig. 3m–o). Based on these experiments, a precise correlation between the analysed proteins and the cellular compartments is not possible. However, irrespective of the performed fixation procedure, it is obvious that hVDAC1 is co-localised in HeLa cells with both Tctex-1 and PBP74.
3.3. Expression of hVDAC1 and partner proteins
In PLB experiments hVDAC1 presented its characteristic voltage dependency. Inserted VDAC1 molecules show the highest relative conductance at applied voltages around 0 mV. Increasing the potential leads to a partial closing of the channel. This is reflected in a bell-shaped curve obtained from the relative conductance as a function of the applied voltage (Fig. 4 closed squares). Applying recombinant partner proteins, the voltage dependency of the channel was significantly modulated. The voltage dependency of hVDAC1 was increased identically by pre-incubated recombinant Tctex-1 (Fig. 4 open triangle) or by addition of Tctex-1 after saturated membrane insertion of hVDAC1 (data not shown). This indicated that the protein–protein binding leads to a channel complex that exhibited lower relative conductance than wild type hVDAC1 when the applied voltage was increased in positive or negative directions. PLB experiments on hVDAC1 pre-incubated with the recombinant PBP74 (Fig. 4 open circles) indicate a strong decrease of voltage dependency. In addition, the recombinant PBP74 (including only a C-terminal segment of PBP74 that was found to interact with hVDAC1 in vivo) affected the channel properties of hVDAC1 in a similar way (Fig. 4 open squares). The voltage dependency of membrane-inserted hVDAC1 was not altered by recombinant PBP74. Consequently, only binding of the PBP74 to hVDAC1 before its membrane insertion favours a more open state of the channel at higher positive or negative applied potentials. Since each recombinant protein alone failed to alter the electrophysiological properties of the PLB (data not shown), changes in channel properties of hVDAC1 are due to specific modulations of hVDAC1’s electrophysiological properties, which result from specific protein–protein interactions.
The distribution of hVDAC1, Tctex-1, and PBP74 was analysed in HeLa cells. As seen on fluorescent microscopy images obtained with specific primary antibodies against each protein and FITC-, or TRITC-conjugated secondary antibodies, the corresponding proteins were labelled in regions between the nucleus and the cell membrane. Fig. 3a and b demonstrate the expression of hVDAC1 and Tctex-1 on formaldehyde-fixed HeLa cells. As seen in Fig. 3c, both proteins could clearly be co-localised. The antibodies applied on these cells did not stain the nuclear regions (Fig. 3d). To preclude cross-reactivity of anti-hVDAC1 antibodies, HeLa cells were transfected with plasmids coding for EGFP–hVDAC1 fusion proteins prior to acetone fixation, and were labelled with anti Tctex-1 antibodies. Fig. 3e and f show the expression of EGFP–hVDAC1 and Tctex-1. The double-staining in Fig. 3 confirms the co-localisation of hVDAC1 with the dynein light chain on HeLa cells. Nuclear staining of this cell is seen in Fig. 3h. A corresponding analysis was performed for hVDAC1 and the PBP74. Antibody-staining specific for hVDAC1 and PBP74 point to a co-localisation in HeLa cells as well (Fig. 3i–l). Additionally, EGFP–hVDAC1 fusion proteins show an overlapping distribution with PBP74 in cytosolic regions of the transfected cells (Fig. 3m–p). Differences between formaldehyde- and acetonefixed cells are worthy of attention. hVDAC1, Tctex-1, and the PBP74 could be detected on formaldehyde-fixed cells throughout the cytoplasma from the cell membrane to the nucleus (Fig. 3a–c and Fig. 3i–k). In contrast, the accumulation of these proteins is visible in areas closer to the cell nucleus on
3.4. Influence on the electrophysiological properties of hVDAC1
1066
C. Schwarzer et al. / The International Journal of Biochemistry & Cell Biology 34 (2002) 1059–1070
Fig. 3. Distribution of hVDAC1 and the interacting proteins in HeLa cells using fluorescent microscopy. Formaldehyde-fixed cells were used for double-immunofluorescence (a–d, and i–l). HeLa cells transfected with EGFP–hVDAC1 plasmids were fixed in acetone prior immunofluorecsence (e-h, and m-p). a: hVDAC1, (b): Tctex-1, (c): hVDAC1 and Tctex-1; (e): EGFP–hVDAC1, (f): Tctex-1, (g): EGFP–hVDAC1 and Tctex-1; (i): hVDAC1, (j): PBP74, (k): hVDAC1 and PBP74; (m): EGFP–hVDAC1, (n): PBP74, o: EGFP-VDAC1 and PBP74. Nuclear staining is seen in (d), (h), (l), and (p). Calibration bars = 5 m.
C. Schwarzer et al. / The International Journal of Biochemistry & Cell Biology 34 (2002) 1059–1070
Fig. 4. Voltage dependence of hVDAC1 in the presence or absence of interacting proteins. The diagram shows the ratio of the conductance G at a given membrane potential, divided by the conductance G0 at +10 mV as a function of the applied membrane potential V (mV). The aqueous phase contained Ringer solution. The membrane consisted of l-␣-diphytanoyl phosphatidylcholine. (hVDAC1 solid squares, n = 21; hVDAC1 + rTctex-1 open triangles, n = 21; hVDAC1 + rPBP74 open circles, n = 14 and hVDAC1 + rPBP74 open squares, n = 33, data are given as means ± S.D.).
4. Discussion The identification of protein–protein interactions is a promising method for expanding our knowledge about proteins. In vivo assays enable the screening of binding proteins under near-physiological conditions in an easy manner. The detergents or denaturing substances often used to isolate membrane proteins in in vitro experiments (and which are known to modulate their natural properties) are not necessary. In our approach, hVDAC1 was used as a bait protein in the CytoTrapTM two-hybrid system. This system permits post-translational modifications of the expressed proteins and allows for the screening of interactions at the cell membrane of yeast cells. It is expected to produce false positive interactions for proteins which are naturally expressed at the cell membrane. Interestingly, the pSos-hVDAC1 bait protein alone was unable to activate the system via the triggering of the Ras signal pathway. This can be explained by the N-terminal fusion of the hSos protein with hVDAC1. We recently observed that hVDAC1, fused with the EGFP at its C-terminus, could be localised at the cell membrane
1067
when heterologously expressed in Xenopus oocytes, whereas N-terminal EGFP–hVDAC1 fusion proteins could not [11]. Modification of the N-terminus inhibits the expression of hVDAC1 at the cell membrane, but enables the use of pSos-hVDAC1 as a bait protein in the applied two-hybrid screening. Interactions of hVDAC1 with both Tctex-1 and PBP74 were identified in vivo using the CytoTrapTM two hybrid system (Fig. 1), and confirmed by overlay assays in vitro (Fig. 2a–c). Additional validation is based on the specific modulation of the channel properties of hVDAC1 by Tctex-1 and PBP74, as observed in artificial PLB experiments (Fig. 4). Immunofluorescence microscopy shows the co-localisation of hVDAC1 with both proteins in HeLa cells (Fig. 3). Tctex-1 is a 14 kDa dynein light chain [19]. As a component of the dynein motor complex, Tctex-1 is involved in the transport of membranous organelles and protein complexes along microtubules [20]. The anterograde transport of mitochondria is driven by the plus end-directed microtubule motor kinesin. It is unclear whether the dynein motor complex is accountable for the retrograde transport of the mitochondria [21]. The identified interaction of the outer mitochondrial membrane protein hVDAC1 and the dynein subunit Tctex-1 suggests a possible relevance for the orientation of mitochondria along microtubules. Additional evidence for a correlation between the dynein motor complex and mitochondria is presented in experiments performed by Caggese et al. [22]. Screening a Drosophila cDNA expression library with antibodies against mitochondrial proteins, the authors found Tctex-1 to be associated with mitochondria. Further evidence for the correlation between VDAC1 and microtubules was reported by the binding of VDAC1 to the microtubules-associated protein MAP2 [23]. These data suggest the involvement of VDAC1 in the orientation of the mitochondria along microtubules. Apart from this mitochondria-based correlation, Tctex-1 participates in the minus end-directed microtubule transport of Rhodopsin [24], Doc2 [25], and CD5 [26]. Correspondingly, the extramitochondrial hVDAC1 could be transported in vesicles along the microtubules by the dynein motor complex. This hypothesis is supported by the expression of VDAC1 in endosomes, which are oriented towards the microtubules minus ends [7,27].
1068
C. Schwarzer et al. / The International Journal of Biochemistry & Cell Biology 34 (2002) 1059–1070
Accordingly, plasmalemmal VDAC1 would be internalised from the cell membrane to the cell interior by the dynein complex. The intracellular trafficking of VDAC1 to the outer mitochondrial membrane, the cell membrane, the sarcoplasmatic reticulum, and endosomes is under debate [10,11,28]. Interactions of VDAC1 with the microtubule-associated proteins Tctex-1 and MAP2 [23], as well as G-actin [29] and the actin-associated protein gelsolin [30] support the idea of a vesicular transport or orientation along the cytoskeleton. The vesicular transport of mitochondrial proteins to extramitochondrial compartments is discussed by Soltys and Gupta [31]. The other protein found to interact with hVDAC1 is PBP74/mtHSP70/GRP75/mortalin (hereafter termed PBP74). The various names describe the putative functions of the protein, as obtained through various experimental approaches. It is known to be involved in antigen processing in mice (PBP74), the mitochondrial protein import pathway in humans (mtHSP70), the pathway regulating glucose response in rats (glucose regulated protein 75, GRP75), and in cell mortality in mice (mortalin) [32–35]. It belongs to the HSP70 family, that is involved in protein folding and protein assembly into multiprotein complexes, and in their subsequent transport across various membranes [36,37]. Interaction of hVDAC1 with the heat-shock protein in mitochondria seems unlikely, since both proteins are expressed at distinct locations, i.e. the outer mitochondrial membrane and the matrix. Apart from this, immuno-microscopical data indicate a localisation of the PBP74 in cytoplasmatic vesicles, ER, endocytotic vesicles or endosomes, and at the cell membrane [32,38]. Since VDAC1 has also been identified in the sarcoplasmatic reticulum, in endosomes, and at the cell membrane, an interaction in these compartments is highly conceivable. The PBP74 is involved in the intracellular trafficking of the fibroblast growth factor 1 and the folding and translocation of the IL-1 receptor 1 [39,40]. These data and the ability of the recombinant PBP74 to bind most likely unfolded hVDAC1, suggest a function of the PBP74 with respect to processing and translocation during the biogenesis of the extramitochondrial VDAC1. Tctex1 influences the relative conductance of pre-incubated as well as membrane inserted hVDAC1
in PLB experiments. The specific modulation of the relative conductance of hVDAC1 is of interest, since both the mitochondrial and the plasmalemmal VDAC1 are considered to need regulation of their permeability. Members of the Bcl-2 protein family interact with VDAC1, showing its possible involvement in apoptotic cell death. Binding of the anti-apoptotic protein Bcl-XL leads to a channel closure of VDAC1 [2,41]. The actin-modulating protein gelsolin has corresponding effects on VDAC [30]. Interaction of VDAC with glycerol-, hexo-, and kreatin-kinase, and NADH [42–45] indicate a possible role in mitochondrial metabolism. Tctex-1 increases the voltage dependency of hVDAC1 by reducing its relative conductance when the applied voltage was increased in positive or negative direction. This means a partial closure of the pore. Whether or not the binding of Tctex-1 to the mitochondrial hVDAC1 participates in apoptotic cell death, or modulates the mitochondrial metabolism remains open for future research. In addition, the regulation of the channel is of special interest for plasmalemmal VDAC1. While plasmalemmal VDAC1 forms large conductance anion channels as observed in patch-clamp experiments in the excised mode, channel activity was monitored after dissociation of a regulating component [10,12,13]. VDAC1 is discussed to be part of the outwardly rectifying chloride channel (ORCC) that is affected in cystic fibrosis [46], and to be involved in cell volume regulation [14,15]. A strict regulation of the channel in this compartment is of special interest because of the channel characteristics discovered for isolated VDAC1 in PLB experiments. Tctex1 incresases the voltage dependency of hVDAC1 and is, therefore, an interesting candidate for the regulation of the channel at the cell membrane. Further studies will indicate how Tctex-1 and PBP74 participates in the channel regulation and/or the cellular trafficing of VDAC1.
Acknowledgements We are grateful to Margret Praetor and Rolf Merker, Department of Immunochemistry, MPI for Experimental Medicine, Göttingen, for performing oligonucleotide synthesis, DNA-sequencing, and digitalisation of the images.
C. Schwarzer et al. / The International Journal of Biochemistry & Cell Biology 34 (2002) 1059–1070
References [1] M. Colombini, Anion channels in the mitochondrial outer-membrane, Curr. Topics Membr. 42 (1994) 73–101. [2] S. Shimizu, M. Narita, Y. Tsujimoto, Bcl-2 family proteins regulate the release of apoptogenic cytochrome c by the mitochondrial channel VDAC, Nature 399 (1999) 483–487. [3] S. Shimizu, Y. Matsuoka, Y. Shinohara, Y. Yoneda, Y. Tsujimoto, Essential role of voltage-dependent anion channel in various forms of apoptosis in mammalian cells, J. Cell. Biol. 152 (2001) 237–250. [4] F.P. Thinnes, H. Götz, H. Kayser, R. Benz, W.E. Schmidt, H.D. Kratzin, N. Hilschmann, To the knowledge of human porins. Part 1. Purification of a porin from human, B-lymphocytes (Porin 31 HL) and the topochemical proof of its expression on the plasmalemma of the parentage cell-line, Biol. Chem. H.-S. 370 (1989) 1253–1264. [5] L. Jürgens, J. Kleineke, D. Brdiczka, F.P. Thinnes, N. Hilschmann, Localization of type-1 porin channel (VDAC) in the sarcoplasmatic reticulum, Biol. Chem. H.-S. 376 (1995) 685–689. [6] V. ShoshanBarmatz, N. Hadad, W. Feng, I. Shafir, I. Orr, M. Varsanyi, L.M.G. Heilmeyer, VDAC/porin is present in sarcoplasmic reticulum from skeletal muscle, FEBS. Lett. 386 (1996) 205–210. [7] S. Reymann, W. Haase, W. Krick, G. Burckhardt, F.P. Thinnes, Endosomes: another extra-mitochondrial location of type-1 porin voltage-dependent anion-selective channels (VDAC), Pflügers Arch. Eur. J. Phys. 436 (1998) 478–480. [8] W.H. Yu, W. Wolfgang, M. Forte, Subcellular-localization of human voltage-dependent anion channel isoforms, J. Biol. Chem. 270 (1995) 13998–14006. [9] G. Bathori, I. Parolini, F. Tombola, I. Szabo, A. Messina, M. Oliva, V. De Pinto, M. Lisanti, M. Sargiacomo, M. Zoratti, Porin is present in the plasma membrane where it is concentrated in caveolae and caveolae-related domains, J. Biol. Chem. 274 (1999) 29607–29612. [10] R. Buettner, G. Papoutsoglou, E. Scemes, D.C. Spray, R. Dermietzel, Evidence for secretory pathway localization of a voltage-dependent anion channel isoform, Proc. Natl. Acad. Sci. U.S.A. 97 (2000) 3201–3206. [11] C. Schwarzer, S. Becker, L.A. Awni, T. Cole, R. Merker, S. Barnikol-Watanabe, F.P. Thinnes, N. Hilschmann, Human voltage-dependent anion-selective channel expressed in the plasmalemma of Xenopus laevis oocytes, Int. J. Biochem. Cell B 32 (2000) 1075–1084. [12] R. Dermietzel, T.K. Hwang, R. Buettner, A. Hofer, E. Dotzler, M. Kremer, R. Deutzmann, F.P. Thinnes, G.I. Fishman, D.C. Spray, D. Siemen, Cloning and in-situ localization of a brain-derived porin that constitutes a large-conductance anion channel in astrocytic plasma-membranes, Proc. Natl. Acad. Sci. U.S.A. 91 (1994) 499–503. [13] U. Janisch, G. Skofitsch, F.T. Thinnes, W.F. Graier, K. Groschner, Evidence for the involvement of plasma membrane porin in agonist-induced activation of vascular endothelial cells, Naunyn-Schmiedeberg’s Arch, Pharmacol. 347 (1993) R72.
1069
[14] F.P. Thinnes, K.P. Hellmann, T. Hellmann, R. Merker, C. Schwarzer, G. Walter, H. Götz, N. Hilschmann, Studies on human porin. Part XXI. Gadolinium opens up cell membrane standing porin channels making way for the osmolytes chloride or taurine: a putative approach to activate the alternate chloride channel in cystic fibrosis, Mol. Genet. Metab. 69 (2000a) 240–251. [15] F.P. Thinnes, K.P. Hellmann, T. Hellmann, R. Merker, U. Brockhaus-Pruchniewicz, C. Schwarzer, G. Walter, H. Götz, N. Hilschmann, Studies on human porin. Part XXII. Cell mem-brane integrated human porin channels are involved in regulatory volume decrease (RVD) of HeLa cells, Mol. Genet. Metab. 69 (2000b) 331–337. [16] A. Aronheim, E. Zandi, H. Hennemann, S.J. Elledge, M. Karin, Isolation of an AP-1 repressor by a novel method for detecting protein–protein interactions, Mol. Cell. Biol. 17 (1997) 3094–3102. [17] K.L. Guan, J.E. Dixon, Eukaryotic proteins expressed in Escherichia coli: an improved thrombin cleavage and purification procedure of fusion proteins with glutathioneS-transferase, Anal. Biochem. 192 (1991) 262–267. [18] R. Benz, E. Maier, F.P. Thinnes, H. Götz, N. Hilschmann, Studies on human porin. Part VII. The channel properties of the human B lymphocyte membrane-derived Porin 31HL are similar to those of mitochondrial porins, Biol. Chem. H.-S. 373 (1992) 295–303. [19] S.M. King, J.F. Dillman, S.E. Benashski, R.J. Lye, R.S. PatelKing, K.K. Pfister, The mouse t-complex-encoded protein Tctex-1 is a light chain of brain cytoplasmic dynein, J. Biol. Chem. 271 (1996) 32281–32287. [20] J. Lane, V. Allan, Microtubule-based membrane movement, BBA Rev. Biomembr. 1376 (1998) 27–55. [21] J.K. Burkhardt, C.J. Echeverri, T. Nilsson, R.B. Vallee, Overexpression of the dynamitin (p50) subunit of the dynactin complex disrupts dynein-dependent maintenance of membrane organelle distribution, J. Cell. Biol. 139 (1997) 469– 484. [22] C. Caggese, G. Ragone, B. Perrini, R. Moschetti, V. De Pinto, R. Caizzi, P. Barsanti, Identification of nuclear genes encoding mitochondrial proteins: isolation of a collection of D. melanogaster cDNAs homologous to sequences in the Human Gene Index database, Mol. Gen. Genet. 261 (1999) 64–70. [23] M. Linden, G. Karlsson, Identification of porin as a binding site for MAP2, Biochem. Biophys. Res. Commun. 218 (1996) 833–836. [24] A.W. Tai, J.Z. Chuang, C. Bode, U. Wolfrum, C.H. Sung, Rhodopsin’s carboxy-terminal cytoplasmic tail acts as a membrane receptor for cytoplasmic dynein by binding to the dynein light chain Tctex-1, Cell 97 (1999) 877–887. [25] F. Nagano, S. Orita, T. Sasaki, A. Naito, G. Sakaguchi, M. Maeda, T. Watanabe, E. Kominami, Y. Uchiyama, Y. Takai, Interaction of Doc2 with tctex-1, a light chain of cytoplasmic dynein: implication in dynein-dependent vesicle transport, J. Biol. Chem. 273 (1998) 30065–30068. [26] A. Bauch, K.S. Campbell, M. Reth, Interaction of the CD5 cytoplasmic domain with the Ca2+ /calmodulin-dependent kinase II delta, Eur. J. Immunol. 28 (1998) 2167–2177.
1070
C. Schwarzer et al. / The International Journal of Biochemistry & Cell Biology 34 (2002) 1059–1070
[27] T. Ichikawa, M. Yamada, D. Homma, R.J. Cherry, I.E.G. Morrison, S. Kawato, Digital fluorescence imaging of trafficking of endosomes containing low-density lipoprotein in brain astroglial cells, Biochem. Biophys. Res. Commun. 269 (2000) 25–30. [28] G. Bathori, I. Parolini, I. Szabo, F. Tombola, A. Messina, M. Oliva, M. Sargiacomo, V. De Pinto, M. Zoratti, Extramitochondrial porin: facts and hypotheses, J. Bioenerg. Biomembr. 32 (2000) 79–89. [29] X. Xu, J.G. Forbes, M. Colombini, Actin modulates the gating of Neurospora crassa VDAC, J. Membr. Biol. 180 (2001) 73–81. [30] H. Kusano, S. Shimizu, R.C. Koya, H. Fujita, S. Kamada, N. Kuzumaki, Y. Tsujimoto, Human gelsolin prevents apoptosis by inhibiting apoptotic mitochondrial changes via closing VDAC, Oncogene 19 (2000) 4807–4814. [31] B.J. Soltys, R.S. Gupta, Mitochondrial proteins at unexpected cellular locations: export of proteins from mitochondria from an evolutionary perspective, Int. Rev. Cytol. 194 (2000) 133– 196. [32] S.Z. Domanico, D.C. Denagel, J.N. Dahlseid, J.M. Green, S.K. Pierce, Cloning of the gene encoding peptide-binding protein 74 shows that it is a new member of the heat-shock protein 70 family, Mol. Cell. Biol. 13 (1993) 3598–3610. [33] T. Bhattacharyya, A.N. Karnezis, A.P. Murphy, T. Hoang, B.C. Freeman, B. Phillips, R.I. Morimoto, Cloning and subcellular-localization of human mitochondrial HSP70, J. Biol. Chem. 270 (1995) 1705–1710. [34] S.M. Massa, F.M. Longo, J. Zuo, S. Wang, J. Chen, F.R. Sharp, Cloning of rat GRP75, an HSP70-family member, and its expression in normal and ischemic brain, J. Neurosci. Res. 40 (1995) 807–819. [35] R. Wadhwa, S.C. Kaul, Y. Sugimoto, Y. Mitsui, Induction of cellular senescence by transfection of cytosolic mortalin cdna in NIH-3T3-cells, J. Biol. Chem. 268 (1993) 22239–22242. [36] F.U. Hartl, J. Martin, W. Neupert, Protein folding in the cell: the role of molecular chaperones HSP70 and HSP60, Annu. Rev. Biophys. Biomol. Struct. 21 (1992) 293–322.
[37] J. G Kiang, G.C. Tsokos, Heat-shock protein 70 kDa: molecular biology, biochemistry, and physiology, Pharmacol. Therap. 80 (1998) 183–201. [38] A.M. Vanbuskirk, D.C. Denagel, L.E. Guagliardi, F.M. Brodsky, S.K. Pierce, A role for heat-shock proteins in antigen processing and presentation, Curr. Topics Microbiol. 167 (1991) 83–92. [39] E. Mizukoshi, M. Suzuki, A. Loupatov, T. Uruno, H. Hayashi, T. Misono, S.C. Kaul, R. Wadhwa, T. Imamura, Fibroblast growth factor-1 interacts with the glucose-regulated protein GRP75/mortalin, Biochem. J. 343 (1999) 461–466. Part 2. [40] G. Sacht, R. Brigelius-Flohe, M. Kiess, H. Sztajer, L. Flohe, ATP-sensitive association of mortalin with the IL-1 receptor type I, Biofactors 9 (1999) 49–60. [41] S. Shimizu, T. Ide, T. Yanagida, Y. Tsujimoto, Electrophysiological study of a novel large pore formed by Bax and the voltage-dependent anion channel that is permeable to cytochrome c, J. Biol. Chem. 275 (2000) 12321–12325. [42] V. Adams, L. Griffin, J. Towbin, B. Gelb, K. Worley, E.R.B. McCabe, Porin interaction with hexokinase and glycerol kinase: metabolic microcompartmentation at the outer mitochondrial-membrane, Biochem. Med. Metab. Biol. 45 (1991) 271–291. [43] E.R.B. McCabe, Microcompartmentation of energy-metabolism at the outer mitochondrial-membrane: role in diabetesmellitus and other diseases, J. Bioenerg. Biomembr. 26 (1994) 317–325. [44] M. Kottke, V. Adam, I. Riesinger, G. Bremm, W. Bosch, D. Brdiczka, G. Sandri, E. Panfili, Mitochondrial boundary membrane contact sites in brain: points of hexokinase and creatine-kinase location, and control of Ca2+ transport, Biochim. Biophys. Acta 935 (1988) 87–102. [45] M. Zizi, M. Forte, E. Blachlydyson, M. Colombini, NADH regulates the gating of VDAC, the mitochondrial outermembrane channel, J. Biol. Chem. 269 (1994) 1614–1616. [46] F.P. Thinnes, S. Reymann, New findings concerning vertebrate porin, Naturwissenschaften 84 (1997) 480–498.