Nuclear transport of protein TTC4 depends on the cell cycle
- First Online:
- Cite this article as:
- Dmitriev, R.I., Okkelman, I.A., Abdulin, R.A. et al. Cell Tissue Res (2009) 336: 521. doi:10.1007/s00441-009-0785-y
- 143 Views
TTC4 (tetratricopeptide repeat domain protein 4) is a putative tumor suppressor involved in the transformation of melanocytes. At present, the relationships between TTC4 and DNA replication proteins are largely unknown, as are the tissue distribution and subcellular localization of TTC4. Using reverse transcription with the polymerase chain reaction, we have observed that the murine TTC4 gene is ubiquitously expressed. Analysis of the TTC4 subcellular localization has shown that, upon overexpression, TTC4 localizes to the cytoplasm. Interestingly, co-expression with a known protein interaction partner, hampin/MSL1, results in the nuclear translocation of the TTC4 protein. The subcellular localization of endogenous TTC4 depends, however, on the cell cycle: it is mostly nuclear in the G1 and S phases and is evenly distributed between the nucleus and cytoplasm in G2. The nuclear transport of TTC4 is apparently a complex process dependent on interactions with other proteins during the progression of the cell cycle. Thus, the dynamic character of the nuclear accumulation of TTC4 might be a potential link with regard to its function in tumor suppression.
The structural family of tetratricopeptide domain proteins (TPR) comprises about one hundred members characterized by repeats of a loose 34-amino-acid consensus and low overall sequence homology. Cellular functions of many of these proteins remain elusive (Blatch and Lässle 1999; Smith 2004). One member, tetratricopeptide repeat domain protein 4 (TTC4), is especially interesting since its gene is frequently mutated in breast cancer (Su et al. 1999, 2000) and perhaps also in melanoma (Poetsch et al. 2000; Irwin et al. 2002). Additionally, TTC4 protein level has been found to be elevated in several malignant melanoma cell lines (Crevel et al. 2008).
Interestingly, the Drosophila ortholog of TTC4, viz., Dpit47, has been shown to form complexes with heat shock proteins HSP90/70 and DNA polymerase (Crevel et al. 2001). This indicates a novel aspect of the regulation of DNA replication under normal conditions and under heat shock. Moreover, TTC4 and Dpit47 interact with another protein involved in DNA replication, viz., CDC6, and this property is conserved from Drosophila to humans (Crevel et al. 2008). TTC4 protein interactions might be much more complex, because TTC4 also interacts with hampin/MSL1, a component of the MYST1 histone acetyltransferase complex, which in turn has at least five components (Dmitriev et al. 2007). Both MYST1 and CDC6 proteins are involved in tumorigenesis: CDC6 has proto-oncogenic activity (Borlado and Méndez 2008), and the downregulation of MYST1 is characteristic for medulloblastomas (Pfister et al. 2008).
Thus, at least one of the functions of TTC4 might be tumor suppression, possibly accomplished through intricate protein-protein interactions. In this paper, we report the subcellular localization of TTC4 and show that its pattern dynamically changes during the progression of the cell cycle, and that the nuclear transport of TTC4 is promoted by hampin/MSL1.
Materials and methods
Mouse NIH-3T3 fibroblasts, rat prostate cells YPEN-1, and HEK-293 cells were from ATCC (LGC Standards, Teddington, Middlesex, UK). Macrophage-like WEHI-3 cells were kindly provided by R.V. Komaleva (Shemyakin-Ovchinnikov Institute of Bioorganic Chemistry, Moscow, Russia). All cell lines were cultured in Dulbecco’s modified Eagle’s medium supplemented with 10% fetal calf serum.
Rabbit anti-TTC4 antibodies were as previously described (Dmitriev et al. 2007). The anti-5′-bromo-2-deoxyuridine (BrdU) monoclonal antibody (BU33) was purchased from Sigma-Aldrich (St Louis, Mo., USA).
Analysis by reverse transcription with polymerase chain reaction
For analysis by reverse transcription with polymerase chain reaction (RT-PCR), total cellular RNA was isolated from murine tissues and cell lines and processed as described previously (Dmitriev et al. 2005). Oligonucleotide primers specific for TTC4 and D-glyceraldehyde-3-phosphate dehydrogenase cDNA were as described earlier (Dmitriev et al. 2005, 2007).
DNA constructs and transfection
Green fluorescent protein-TTC4 (GFP-TTC4) and TTC4-DsRed fluorescent protein (TTC4-RFP) chimeras were obtained by cloning the PCR-amplified mouse TTC4 open reading frame (ORF) into pEGFP-C1 and pDsRed-express-N1 vectors (Clontech, Mountain View, Calif., USA), respectively. Hampin B-GFP chimera was obtained by cloning PCR-amplified mouse hampin B ORF into pEGFP-N3 vector (Clontech), Plasmid DNA “hampin C-GFP” was as described earlier (Dmitriev et al. 2006). GFP-DNA-methyl transferase I (GFP-DNMT1) and RFP-Ligase I constructs were kindly provided by M.C. Cardoso (Max Delbrück Center for Molecular Medicine, Berlin, Germany). Transfections were performed with liposome reagents: Lipofectamine 2000 (Invitrogen, Carlsbad, Calif., USA) or a 1:1 mixture of a polycationic cholesterol derivative, a proprietary lipid mixture kindly provided by Prof. G.A. Serebrennikova (M.V. Lomonosov Academy of Fine Chemical Technology, Moscow, Russia).
Immunocytochemistry and Western blotting
The procedures for immunofluorescence microscopy were performed as described earlier (Dmitriev et al. 2007). Briefly, cells grown on coverslips were fixed in methanol (−20°C, 40 min) or in 4% formaldehyde in phosphate-buffered saline (20°C, 10 min). Formaldehyde-fixed cells were permeabilized with 0.25% Triton X-100. For BrdU labeling, cells were treated with DNase I (Easwaran et al. 2005). Fixed cells were blocked with 5% bovine serum, incubated with primary antibodies followed by Alexa-Fluor-conjugated secondary antibodies (Invitrogen) and mounted in a 4,6-diamidino-2-phenylindole (DAPI)-containing solution before imaging on a Nikon Diaphot 300 fluorescent microscope. For actin cytoskeleton labeling, Alexa-Fluor-546 conjugated to phalloidin (Invitrogen) was used according to the recommendations of the manufacturer. Western blotting was carried out by using ECL Plus reagent (GE Healthcare, Giles, UK) as described (Pestov et al. 2002).
Synchronization and determination of cell-cycle phase
Serum starvation was chosen as the method for the synchronization of cultured cells (Vriz et al. 1992). Briefly, 3T3 fibroblasts were incubated in serum-free medium for 70 h to ensure their entry into G0. After the addition of serum, an incubation of approximately 7 h was required for their entry into G1 phase, and of 12 h for entry into S phase (Vriz et al. 1992). Pulse-labeling (15 min) with BrdU (100 μM) was used to detect entry into S phase followed by immunostaining with anti-BrdU antibodies. Stages of the cell cycle in cells transfected with GFP-DNMT1 and RFP-Ligase I were determined by using criteria described by Easwaran et al. (2005).
Cells transfected with GFP-TTC4 were incubated at +42°C for 2 h, fixed in 4% formaldehyde, and processed for fluorescence imaging as described above.
Results and discussion
The observation that hampin/MSL1 interacts with TTC4 both in vitro and in vivo (Dmitriev et al. 2007) prompted us to investigate this interaction in more detail. We constructed plasmids encoding chimeras of TTC4 with fluorescent proteins for the transfection of 3T3 fibroblasts. Unexpectedly, both GFP-TTC4 (N-terminal fusion) and TTC4-RFP (C-terminal fusion) chimeras were found to be localized in the cytoplasm, with no significant nuclear accumulation (Figs. 1b–e). To verify that transfection with plasmids encoding the TTC4 fluorescent chimera resulted in its overexpression, we stained transfected cells with anti-TTC4 antibodies. (Figs. 1f, g). As expected, transfected cells showed much brighter staining than untransfected cells. In addition, we checked the location of the TTC4 protein expressed with no fluorescent tags and obtained similar results (not shown).
The cytoplasmic location of protein TTC4 is an interesting phenomenon, because many tumor suppressors are affected by dysregulated nucleocytoplasmic transport (Fabbro and Henderson 2003). Therefore, we decided to study the subcellular localization of the endogenous TTC4 in more detail. In asynchronous cultures, staining with anti-TTC4 antibodies was found to be variable between individual cells: from predominantly nuclear to diffuse whole-cell distributions (see Electronic Supplementary Material, Fig. S1). Importantly, similar results were obtained with several different cell lines: 3T3 fibroblasts, macrophage-like WEHI-3, rat YPEN-1, and human HEK-293 cells (results not shown). The heterogeneity observed within each cell line may be a result of differences between cell-cycle phases in cells.
To test this possibility, we studied the localization of the protein TTC4 in 3T3 cells synchronized by serum starvation (Vriz et al. 1992). We observed that, after prolonged incubation without serum (cells in G0 phase), TTC4 was distributed evenly between the nucleus and cytoplasm (see Electronic Supplementary Material, Fig. S2). After the addition of serum, TTC4 was translocated to the nucleus: mostly nuclear staining could be observed after 7 h (cells in G1 phase). Later on, cytoplasmic labeling became stronger and, after a 23-h incubation, was of equal density to nuclear labeling.
In addition, we used a system of cell-cycle markers based on fluorescent chimeras of DNMT1 and DNA-Ligase I (Easwaran et al. 2005). In the S phase, DNA-Ligase I has a punctate pattern that co-localized with DNA replication foci and with BrdU. In non-S phases (G1, G2, M), DNA-Ligase had a diffuse nucleoplasmic location, whereas DNMT1 was localized on centromeric chromatin at S and G2 phases. The merged images of DNMT1 and DNA-Ligase fluorescence allowed the clear identification of G1, S, and G2 phases (Easwaran et al. 2005).
By using this system, TTC4 was found to be predominantly nuclear in G1 phase (Figs. 3d–g’’’), with no significant differences between G1 and S sub-phases. In identified G2 cells, however, the TTC4 localization shifted to the cytoplasm. Thus, TTC4 is nuclear during G1 and S phases, whereas G2 and M phases are marked by the existence of an abundant cytoplasmic pool of TTC4.
This dynamic pattern of localization may be a result of the regulation of the protein level of TTC4 at the various stages of cell cycle, as has been shown for many tumor suppressors (Delmolino et al. 2001; Orlando et al. 2008). Therefore, we checked the relative quantity of the TTC4 in synchronized cells at different time intervals after serum addition by using Western blotting and found that the TTC4 level remained constant (Fig. 3c).
Since TTC4 is a co-chaperone of HSP90/70 (Crevel et al. 2001, 2008), its function may be affected by heat shock. However, we have not detected any remarkable effects of heat shock on the subcellular location of endogenous TTC4 in 3T3 fibroblasts (results not shown).
Here, endogenous TTC4 has been shown to have a dynamic subcellular localization that depends on the cell cycle. This is consistent with a recently found link between TTC4 and DNA replication enzymes (Crevel et al. 2008), such as DNA primase or CDC6. Indeed, both of these enzymes also exhibit dynamic patterns of subcellular localization. CDC6 is nuclear in G1 and cytoplasmic in S phase (Delmolino et al. 2001). Moreover, the lack of nuclear localization of the overexpressed TTC4 is similar to the results obtained on DNA polymerase (p180) by Mizuno and co-workers (Miyazawa et al. 1993; Mizuno et al. 1998). Importantly, GFP-tagged p180 is cytoplasmic when overexpressed alone but moves to the nucleus upon co-expression with the p68 subunit (Mizuno et al. 1998).
Because the Drosophila homolog of TTC4 protein has previously been shown to be a co-factor of p180 (Crevel et al. 2001), we have also tried to co-express murine TTC4 with this protein. Unfortunately, we have found it impossible or difficult to obtain cells overexpressing proteins p180-GFP and TTC4-RFP simultaneously, both in the case of nuclear p180 (co-expressed with another DNA polymerase subunit, p68) and in the case of cytoplasmic p180 (data not shown). We hypothesize that the overexpression of both proteins (p180 and TTC4) is highly toxic to cells.
We have shown that cytoplasmic TTC4 is translocated to the nucleus upon co-expression with its nuclear partner hampin/MSL1. The question remains open, however, as to whether the dynamic localization of endogenous TTC4 protein during the cell cycle is governed by hampin alone. This is feasible since H4 K16 acetylation, a process in which hampin is involved (Smith et al. 2005), has been reported to be dependent on the cell cycle (Rice et al. 2002).
The change in the subcellular localization of the protein TTC4 in G2 may be caused by a blockade of nuclear transport, an interaction with a cytoplasmic binding molecule, or by the degradation of its nuclear pool. However, we suggest that this phenomenon occurs mainly because of accelerated nuclear export. Indeed, the dynamic localization of TTC4 is typical of proteins that possess nuclear export sequences (NES; Pemberton and Paschal 2005). At least one active NES, viz., -55IRGALCHLEL16-, can be predicted in the amino acid sequence of TTC4 according to the rules devised by Pemberton and Paschal (2005). We have tried to determine whether TTC4 localization is dependent on CRM1 by treating transfected cells with leptomycin B. No nuclear translocation occurs after this treatment. Therefore, the nuclear export of TTC4 is CRM1-independent (see Electronic Supplementary Material, Fig. S3).
We are grateful to M.C. Cardoso for providing GFP-DNMT1 and RFP-Ligase I plasmid DNA, R.V. Komaleva for the generous gift of the WEHI-3 cell line, Fumio Hanaoka and Takeshi Mizuno for the p68 and p180 expression constructs, E.P. Kopantsev and O.V. Zatsepina for advice regarding cell synchronization procedures, S. Volkov, O. Molokoyedova, and N. Bystrov for the synthesis of oligonucleotides, V. Baklaushev for help with antibodies, M. Maslov and Prof. G.A. Serebrennikova for transfection reagents, and N. Stratienko for providing the HeLa cell line.