- Email: [email protected]
Characterization of the novel human transmembrane protein 9 (TMEM9) that localizes to lysosomes and late endosomes
BBRC Biochemical and Biophysical Research Communications 297 (2002) 912–917 www.academicpress.com
Characterization of the novel human transmembrane protein 9 (TMEM9) that localizes to lysosomes and late endosomesq Marit Kveine, Ellen Tenstad, Guri Døsen, Steinar Funderud, and Edith Rian* Department of Immunology, The Norwegian Radium Hospital, 0310 Oslo, Norway Received 26 August 2002
Abstract We have identified and characterized the novel human transmembrane protein 9 (TMEM9). TMEM9 encodes a 183 amino-acid protein that contains an N-terminal signal peptide, a single transmembrane region, three potential N-glycosylation sites, and three conserved cys-rich domains in the N-terminus, but no hitherto known functional domains. The protein is highly conserved between species from Caenorhabditis elegans to man and belongs to a novel family of transmembrane proteins. The TMEM9 gene consists of at least 6 exons and is localized to chromosome 1q41. TMEM9 mRNA is expressed in a wide range of tissues and cells. COS-1 cells transfected with a TMEM9 expression plasmid gave three bands of about 28, 31, and 33 kDa representing glycosylated forms of TMEM9 with a protein backbone of about 26 kDa. In COS-1 cells transfected with a TMEM9-GFP expression construct ,TMEM9GFP is co-expressed with LAMP1 on late endosomes and lysosomes as well as on ER. Thus, TMEM9 is a phylogenetically conserved, widely expressed transmembrane protein with a potential, but unknown function in intracellular transport. Ó 2002 Elsevier Science (USA). All rights reserved. Keywords: Transmembrane protein type I; Signal sequence trap; Novel protein family; Late endosomes; Lysosomes; Chromosome 1
B lymphopoiesis is a stepwise process consisting of initially, a stroma-dependent phase including adherence to and exchange of factors with the cells in the bone marrow microenvironment, and subsequently, a stromaindependent part where the cells further differentiate and migrate out of the bone marrow to peripheral lymphoid organs [1]. Our main focus is to identify and characterize proteins important for the human stroma-dependent B cell differentiation. To identify genes encoding secreted and membrane expressed proteins in a pro-B cell line, we therefore employed a signal sequence trap (SST) based on retroviral expression (SST-REX) [2]. SSTs are methods used to clone and identify the cDNAs of proteins that are guided to the secretory pathways and cellular membranes by the their N-terminal, hydrophobic signal peptide [3–5]. Thus, not only secretory and plasma membrane-bound factors, but also proteins on internal membranes (ER, Golgi network, mitochondria q
TMEM9 GenBank Accession Number is AY138587. * Corresponding author. Fax: +47-22-500-730. E-mail address: [email protected] (E. Rian).
and organelles of the endocytic pathway), can be identified by a signal sequence trap. However, the latter may also be important for cellular communication, in particular proteins on endosomes and lysosomes that may participate in endocytosis of ligand-bound receptors and receptor degradation or recycling [6]. We report here the cloning and identification by SSTREX, and the subsequent characterization of a novel human transmembrane protein, TMEM9, belonging to a new protein family. The mRNA is widely expressed, including in B cells and other hematopoietic cells. Transfected cells produce three glycosylated forms of TMEM9 and the protein co-localizes with LAMP1 on late endosomes and lysosomes, suggesting a role for TMEM9 in intracellular transport.
Materials and methods Reagents and antibodies. The following antibodies were used: mouse anti-human myc antibody (9E10 hybridoma, [7]), rabbit antimouse IgG-HRP (Dako AS, Glostrup, Denmark), mouse anti-human LAMP1 and EEA1 antibodies (provided by H. Stenmark, Department
0006-291X/02/$ - see front matter Ó 2002 Elsevier Science (USA). All rights reserved. PII: S 0 0 0 6 - 2 9 1 X ( 0 2 ) 0 2 2 2 8 - 3
M. Kveine et al. / Biochemical and Biophysical Research Communications 297 (2002) 912–917 of Biochemistry, The Norwegian Radium Hospital), and Cy3-conjugated goat anti-mouse and goat anti-rabbit antibodies (Amersham Pharmacia Biotech, Uppsala, Sweden). Tunicamycin, saponin, phenylmethylsulfonyl fluoride (PMSF), Tween 20, and aprotinin were supplied from Sigma Chemicals (Steinheim, Germany). Polybrene was purchased from Speciality Media (Phillipsburg, USA) and IL-3 was from R&D systems (Abingdon, GB). Cells. The following human cell lines were used in this study: the pro-B cell lines TOM-1 and BV173 (both provided by Dr. T. Logtenberg, Utrecht, The Netherlands), the pre-B cell lines Reh (ATCC CRL-8286) and Nalm 6 [8], the mature B cell lines Bjab (provided by G. Kline, Basel Institute for Immunology, Basel, Switzerland) and Daudi (ATCC CCL-213), the plasmacytoid cell lines U266 (ATCC TIB 196), the T cell lines Jurkat (ATCC TIB-152) and HPB-ALL (provided by M. Greaves, Leukaemia Research Fund Centre, Institute of Cancer Research, London, United Kingdom), the myeloid cell lines KG1-A (ATCC CCL-246), HL-60 (ATCC CCL-240), and U937 (ATCC CRL-1593), the erythroid precursor cell line K562 (ATCC CCL-243), the human larynx carcinoma HEp-2 (ATCC CCL 23) and COS-1 (ATCC CRL1650). All cell lines were grown in RPMI 1640 medium supplemented with 10% FCS, except COS-1 and HEp-2 cells that were grown in DulbeccosÕs modified EagleÕs medium (DMEM) supplemented with 10% FCS. All the cells were cultured at 37 °C in a humidified atmosphere with 5% CO2 . SST-REX cDNA library construction and signal sequence trap performance. Total RNA was prepared from the human pro-B lymphoid cell line BV173 by standard guanidine thiocyanate methods and polyðAÞþ RNA was further purified by oligo(dT)-beads (Dynal, Oslo, Norway). cDNA was synthesized from 5 lg polyðAÞþ RNA using random hexamers and the Stratagene cDNA synthesis kit (Stratagene, Cedar Creek, USA) according to manufacturerÕs instructions. BstXI adapters (Invitrogen, Carlsbad, CA) were ligated to the blunted cDNA ends before separation through SizeSep 400 Spin Column (Pharmacia, Uppsala, Sweden) and insertion of the cDNA into BstXI sites of the retroviral vector pMX-SST. The cDNA library was electroporated into competent XL1-blue bacteria. Plasmid DNA was prepared using the Quiagen Maxi kit (Qiagen, Hilden, Germany). The SST-REX was performed as described by Kitamura [2] with some modifications. Briefly, the BV173 cDNA library was converted to retrovirus by transfection into the packaging cell line Phoenix Eco (G.P. Nolan, Stanford University, CA, USA) using Fugene transfection agent (Roche, Basel, Switzerland). The supernatant contained 106 viral particles per ml as determined from the titer of the control vector pMX-GFP on NIH3T3 cells in a parallel experiment. 0:3 106 viral particles were used to infect 1:2 106 mycoplasma free Ba/F3 cells (DSMZ, Braunschweig, Germany) by spinning the cells (500g, 1 h) in the presence of Polybrene (4 lg/mL) and IL-3 (3 ng/mL). The resulting infection efficiency in Ba/F3 cells was 10% as assessed from the infection efficiency of a control vector pMX-GFP in Ba/F3 cells. The cells were deprived of IL-3 the following day and transferred to 96-well multititer plates to reveal the IL-3 independent clones. Integrated cDNAs from these Ba/F3 clones were PCR-amplified using vector primers on genomic DNA (QiaAmp DNA Blood Minikit, Qiagen, Hilden, Germany) and sequenced. Cloning of TMEM9 cDNA. One of the novel clones identified by SST-REX showed full homology to a DNA encoding a hypothetical protein clone MGC:891. Analysis of the ORF indicated a transmembrane type I protein and the clone was, therefore, chosen for further characterization. The putative protein coding part including 5 bp upstream of the predicted ATG was PCR amplified from a prostate cDNA library (provided by A. Foss a, [9]) using site specific primers (forward primer 50 -TAG GAA GCT TTA AGC ATG AAG CTC TTA-30 and reverse primer for vector with myc-His epitopes 50 -ATC CTC GAG CGG CCG CCG CTG AGC ATC TTG TGC CGA-30 and reverse primer for the pEGFP-N3 vector 50 -ATC CTC GAG GAT CCG CTG AGC ATC TTG TGC CGA-30 ). The primers included restriction enzyme sites HindIII, NotI, and BamHI, respectively.
913
The TMEM9 cDNA was either ligated in-frame upstream of the mycHis epitope cDNA in pcDNA3.1/myc-His (Invitrogen, Carlsbad, USA) or in-frame upstream of green fluorescent protein (GFP) in pEGFP-N3 (Clontech, Palo Alto, USA). The constructs were used to produce mycHis-tagged TMEM9 and TMEM9-GFP fusion protein, respectively. Northern blot analysis. Total RNA was isolated by the RNeasy mini kit (Qiagen, Hilden, Germany). Twenty lg total RNA from each sample was loaded in each well and subjected to Northern blot analysis as previously described [10]. The membranes were hybridized overnight with a 32 P-dCTP-labeled TMEM9 cDNA probe (550 bp ORF fragment, isolation described above), thereafter stripped and hybridized with a probe against 18S rRNA (GenBank Accession No. AJ271813) to monitor loading differences. Two human multiple tissues Northern blots (Clontech H2 and H3, Palo Alto, USA) were hybridized with the TMEM9 cDNA probe according to the instructions of the manufacturer. Signals were detected by phosphoimager (Molecular Dynamics, Amersham Pharmacia Biotech, UK). Western blot analysis. Plasmids encoding TMEM9 tagged with the myc-His epitopes or vector alone (pcDNA3.1/myc-His) were transfected into COS-1 cells using Lipofectamine (Invitrogen, Carlsbad, USA). One to two days post-transfection, 1 106 cells were lysed with 0.05% saponin in PBS supplemented with 30 lg/ml aprotinin and 1 mM phenylmethylsulfonyl fluoride. For tunicamycin-treatment, the transfected cells were incubated with tunicamycin (5 lg/ml) or vehicle (0.025% DMSO) for 16 h before cell lysis. Near equal amounts of total protein extract (related to cell number) were denatured (475 mM bmercaptoethanol) and separated on a 10% polyacrylamide gel, followed by immunoblotting using mouse anti-myc as primary antibody and rabbit anti-mouse HRP as secondary antibody. The proteins were visualized with the ECL system (Amersham, Uppsala, Sweden). The loading was evaluated by Ponceau red staining of the filters. Confocal fluorescence microscopy. HEp-2 cells were grown on coverslips and transiently transfected with TMEM9-GFP in pEGFP-N3 or pEGFP-N3 as control. The transfection reagent Fugene was used according to producerÕs protocol. One day post-transfection, the cells were fixed with 3% paraformaldehyde and subsequently incubated with antiLAMP1 antibody, followed by a secondary antibody (Cy3-labeled antimouse antibody) in a solution containing PBS and 0.05% saponin as previously described [11]. The cellular localization of TMEM9 was observed by examination of coverslips with a Leica TCS NT confocal scanning microscope system equipped with a Kr/Ar laser and a PL Fluotar 100/1.30 oil immersion objective (Leica Microsystems, Heidelberg, Germany). Appropriate emission filter settings and controls were included to exclude bleed-through effects.
Results and discussion Cloning and structural characterization To identify factors important for human stromadependent B lymphopoiesis, we have employed SST-REX [2] on human pro-B lymphoma cells BV173 as a model system as these cells adhere to bone marrow stroma cells and partly differentiate to IgM-positive cells in stroma coculture (not shown). RNA from this cell line was used to construct a cDNA library in the pMX-SST vector, and a preliminary screening employing the SST-REX method [2] resulted in the identification of 18 different known cDNAs and 4 which represented novel sequences. One of the novel sequences with a potential open reading frame (ORF) with properties of a transmembrane type I protein (see below) was selected for further characterization.
914
M. Kveine et al. / Biochemical and Biophysical Research Communications 297 (2002) 912–917
Fig. 1. (A) Amino-acid sequence alignment of TMEM9 from human (hTMEM9), mouse (mTMEM9), and Tokifugu rubripes (trTMEM9) with the human homologous hypothetical protein hC11ORF15 and its murine orthologous hypothetical protein mD7H11ORF15, as well as with the orthologous hypothetical proteins from C. elegans T16732 and D. melanogaster AAF52042. The alignment was obtained by the use of ClustalW (http:// www.ch.embnet.org/software/ClustalW.html). Black shading indicates identities, gray shading indicates similarities. Potential N-glycosylation sites are marked with dashes. The predicted signal peptide cleavage site is marked with an arrow, and the suggested transmembrane (TM) region is marked with a line. Cystein-rich domains are marked with asterisks. (B) Phylogenetic analysis of hTMEM9 and the related proteins aligned in (A). The dendrogram was obtained by using Protpars in the Phylip program package (http://bioweb.pasteur.fr/seqanal/phylogeny/phylip-uk.html). (C) The genomic structure characterized of at least 6 exons represented by gray boxes (not on scale). The translational start and stop codons are marked.
Homology search performed with the BLAST program [12] (http://www.ncbi.nlm. nih.gov) showed that the cDNA of 313 bp was identical to that of the hypothetical
protein, clone MGC:891 (GenBank Accession No. BC001106, Hs.181444), the uncharacterized DERP4 (Accession No. AB013909), and clone HSPC186
M. Kveine et al. / Biochemical and Biophysical Research Communications 297 (2002) 912–917
identified in a hematopoietic stem and progenitor cell cDNA library [13]. The cDNA included the ATG translation start codon and a 62 bp upstream sequence. Sequence information from clone MGC:891 was used to construct primers and the protein coding part of the cDNA was amplified from a prostate cDNA library. The cDNA encodes an ORF of 183 amino-acid (aa) residues (Fig. 1A). Analysis of the predicted protein sequence using the SignalP, PSORTII, and ScanProsite programs (http://www.expasy.ch) indicated a protein with an Nterminal signal peptide and a single transmembrane spanning region (transmembrane protein type I) (Fig. 1A). Three N-glycosylation sites in the N-terminal part of the deduced mature protein were suggested (dashes, Fig. 1A). The first site, however, is not likely to be glycosylated, since it is predicted to be the first amino acid in the mature protein. Apart from these features, no known protein domains or motifs pointing to a function or cellular localization of this protein were found. The protein was named transmembrane protein 9 (TMEM9, approved by the Human Genome Nomenclature Committee (http:// www.gene. ucl.ac.uk/nomenclature). The sequence data has GenBank entry AY138587. Database homology searches identified one hypothetical human homolog with 57% aa identity (74% positivity) to TMEM9 (C11orf15, Accession No. XP_011973) (Fig. 1A). Additionally, hypothetical orthologs in several species were located; mouse (mTMEM9, 95% aa identity, 96% positivity, Accession No. BAB23903), Takifugu rubripes (trTMEM9, 76% aa identity, 85% positivity, Accession No. T30536), Drosophila melanogaster (31% aa identity, 48% positivity, Accession No. AAF52042) and Caenorhabditis elegans (28% aa identity, 46% positivity, Accession No. T16732). Finally, the human homolog has a hypothetical mouse ortholog (D7H11orf15, Accession No. BAB25033) showing 56% aa identity (73% positivity) to TMEM9 (see later). We note that Zhang et al. [13] have also assigned the putative ORF of clone HSPC186 as a potential transmembrane protein with several orthologs. Alignment of the ORFs by ClustalW demonstrates several regions of very high homology (Fig. 1A). First, the suggested transmembrane region of TMEM9 (approximately aa 90–112) is highly homologous between all hypothetical proteins, indicating that all seven are transmembrane proteins. Similar transmembrane topology is suggested by PSORTII for all proteins, except for the C. elegans version (T16732), which may have three transmembrane spanning regions and an uncleavable N-terminus. Second, the C-terminal, cytoplasmic regions of the seven ORFs are nearly identical, suggesting that this domain is functionally important. Third, three conserved cystein-rich domains are found in the N-terminal regions (marked with asterisk, Fig. 1A). These regions may participate in protein folding, protein interactions, and multimerization. Finally, the suggested N-glycosylation sites are partly conserved, and thus, of
915
potential importance for the protein structure and function. Phylogenetic analysis of the protein sequences reflects the close evolutionary relationship between these proteins (Fig. 1B). The high degree of homology between species indicates that these proteins build a novel family of transmembrane proteins highly conserved throughout the phylogenesis, and with potentially important function(s) in all these organisms. Chromosomal localization and genomic organization BLAST with the TMEM9 cDNA against the human genome database (http://www.ncbi.nlm.nih.gov) shows that the TMEM9 gene is localized to chromosome 1q41 (LOC51235). Sequence information from Homo sapiens chromosome 1 clone RP5-894H24 (Accession No. AL 139159), currently undergoing sequencing, suggests that the gene spans at least a 9 kb sequence and that the genomic structure of the gene is characterized by at least 6 exons (Fig. 1B). The ATG translation start site is found on exon 2, whereas the translation stop codon and a potential ATTAAA polyadenylation signal is on exon 6. mRNA expression analysis Northern blot analysis of human tissue RNA blots shows that a major 1.7 kb transcript and a minor 1.1 kb form of TMEM9 mRNA are expressed in a wide range of tissues (Fig. 2A). The minor transcript may represent an alternatively spliced form. TMEM9 mRNA is abundant in adrenal gland, thyroid gland, ovary, testis, and prostate, and to a lesser degree expressed in trachea, spinal cord, stomach, colon, small intestine, thymus, and spleen (Fig. 2A). Low expression is observed in bone marrow, lymph node, and peripheral blood lymphocytes (PBL). However, a more thorough analysis of cell lines of hematopoietic origin demonstrates TMEM9 mRNA expression in cell lines of B lineage cell origin (Tom-1, BV173, Reh, Nalm6, Bjab, Daudi, and U266), T lineage cell origin (HPB-All, Jurkat, and JM), myeloid origin (KG1a, HL60, and U937), and erythroid origin (K562) (Fig. 2B). There is no obvious correlation between the differentiation level represented by the varying cell lines and the mRNA expression level of TMEM9. Western blot analysis and protein deglycosylation We verified that the TMEM9 ORF results in a translated product by producing TMEM9 with a C-terminal myc-His epitope tag (Fig. 3). Total cell lysates from transfected COS-1 cells were analyzed by Western blotting, demonstrating a band triplet of approximately 28, 31, and 33 kDa (arrows). The three bands most likely represent glycosylated forms of the TMEM9 protein, since several potential N-glycosylation sites are found in the amino-acid sequence. Additionally, a heavier band of
916
M. Kveine et al. / Biochemical and Biophysical Research Communications 297 (2002) 912–917
Fig. 2. Northern blot analysis of TMEM9 mRNA expression. (A) Commercial human tissue blots show expression of two TMEM9 transcripts of approximately 1.7 and 1.1 kb. Molecular weight markers are to the right. (B) Expression of TMEM9 mRNA in hematopoietic cell lines of B lineage cell origin (Tom-1, BV173, Reh, Nalm6, Bjab, Daudi, and U266), T lineage cell origin (HPB-All, Jurkat, and JM), myeloid (KG1a, HL60, and U937), and erythroid origin (K562). Probe against 18S rRNA reveals loading differences. The figure is composed of lanes from a single filter exposure.
approximately 60 kDa was routinely observed. We assume that this band may represent a dimer form of the protein formed by interaction between one or more of the
Fig. 3. Western blot analysis of recombinant, myc-His-tagged TMEM9. COS-1 cells were transfected with a plasmid containing the TMEM9 myc-His construct or empty vector (pcDNA3.1/myc-His). After transfection the cells were treated with tunicamycin (T) or vehicle (C) before cell lysis and Western blot analysis. The arrows point out the three glycosylated forms of TMEM9. Molecular weight markers are to the left.
cystein-rich domains found in the N-terminal part. To test whether the protein bands were glycosylated forms of TMEM9, we treated the transfected cells with tunicamycin (5 lg/ml), which inhibits N-glycosylation in the ER. Tunicamycin-treated cells produced a major, single 26 kDa form of myc-His tagged, TMEM9 as well as a weak band of approximately 52 kDa. The molecular masses correspond to the theoretical Mw of the deduced protein sequence of 20.6 kDa, with an addition of 3.7 kDa representing the tag and a dimer form of the tagged protein, respectively. The different forms of recombinant TMEM9 with the myc-His-tag were also found in HEp-2 cells, and both COS-1 and HEp-2 cells produced a similar pattern of glycosylated TMEM9-GFP fusion protein (data not shown). Thus, these data suggest that TMEM9 exists in three glycosylated forms that may have distinct functional and/or affinity properties. Parts of the glycosylation pattern may result from unprocessed carbohydrate chains that render in the ER due to overloading of the transfected system (see below). Cellular localization Since proteins with N-terminal signal peptides may localize to internal cell membranes, as well as the plasma membrane and to secretory vesicles, it is important to determine the cellular distribution of novel proteins identified by signal sequence trap. PSORTII analysis of the protein sequence gave no clue as to which cellular membrane the protein can be directed to. We, therefore, transfected COS-1 cells with a plasmid containing the TMEM9-GFP construct. The empty pEGFP-N3 plasmid and mock transfected cells were used as negative controls (not shown). Confocal microscopy of transfected cells demonstrated fluorescence staining in ER and cytoplasmic vesicles that co-stained with an antiLAMP1 antibody, which is a marker for late endosomes and lysosomes (Fig. 4). No GFP staining was apparent in the nucleus (Fig. 4). TMEM9-GFP did not co-localize with EEA1, a marker for early endosomes (not shown). This indicates that late endosomes and lysosomes are the cellular localization sites for TMEM9. We assume that the strong fluorescence observed in ER is a result of overproduction and accumulation of the protein in transiently transfected cells. Cells transfected with the control plasmid (pEGFP-N3) showed staining mainly in the nucleus, whereas mock transfected cells and cells treated with secondary antibody alone showed no fluorescence staining (not shown). In conclusion, we have characterized a novel human gene and gene product TMEM9. TMEM9 is a type I transmembrane protein that belongs to a novel and phylogenetically conserved family of transmembrane proteins. The gene, consisting of at least 6 exons, is localized to chromosome 1q41. TMEM9 mRNA is widely expressed and transfected cells produce three glycosylated
M. Kveine et al. / Biochemical and Biophysical Research Communications 297 (2002) 912–917
917
Fig. 4. Confocal fluorescence microscopy of fixed COS-1 cells transfected with the TMEM9-GFP expression plasmid. (A) A cell expressing TMEM9GFP shows staining in ER and vesicular structures. (B) Immunostaining of the same cell with anti-LAMP1 antibody. (C) Merged image where yellow staining shows co-localization of TMEM9-GFP and LAMP1.
forms of a myc-His-tagged TMEM9 protein. TMEM9GFP fusion protein partly co-localizes with LAMP1 on late endosomes and lysosomes. The function of TMEM9 remains to be elucidated, but the cellular localization suggests that TMEM9 may participate in intracellular transport.
Acknowledgments We thank H. Stenmark and C. Raiborg, Department of Biochemistry, The Norwegian Radium Hospital, Norway, for providing the LAMP1 and EEA1 antibodies and for invaluable help with the confocal laser scanning microscopy experiments. We are grateful to T. Kitamura, Department of Hematopoietic Factors, University of Tokyo, Japan, for sharing the pMX-SST plasmid and experience regarding the SST-REX procedure with us, and to G.P. Nolan, Stanford University, CA, USA, for the Phoenix Eco packaging cell line. Thanks to A. Foss a, Department of Immunology, The Norwegian Radium Hospital, for providing the DU145 prostate cDNA library and to E.B. Smeland, Department of Immunology, The Norwegian Radium Hospital, for carefully reading the manuscript.
References [1] T.W. LeBien, Fates of human B-cell precursors, Blood 96 (2000) 9–23. [2] T. Kojima, T. Kitamura, A signal sequence trap based on a constitutively active cytokine receptor, Nat. Biotechnol. 17 (1999) 487–490. [3] E. Rian, T. Wabakken, H. Hauge, H.C. Aasheim, A signal sequence trap based on cell enrichment using anti-CD19 antibody coated magnetic beads, Scand. J. Immunol. 54 (2001) 280–284.
[4] R.D. Klein, Q. Gu, A. Goddard, A. Rosenthal, Selection for genes encoding secreted proteins and receptors, Proc. Natl. Acad. Sci. USA 93 (1996) 7108–7113. [5] K. Tashiro, T. Nakamura, T. Honjo, The signal sequence trap method, Methods Enzymol. 303 (1999) 479–495. [6] S. Mukherjee, R.N. Ghosh, F.R. Maxfield, Endocytosis, Physiol. Rev. 77 (1997) 759–803. [7] G.I. Evan, G.K. Lewis, G. Ramsay, J.M. Bishop, Isolation of monoclonal antibodies specific for human c-myc proto-oncogene product, Mol. Cell Biol. 5 (1985) 3610–3616. [8] R. Hurwitz, J. Hozier, T. LeBien, J. Minowada, K. GajlPeczalska, I. Kubonishi, J. Kersey, Characterization of a leukemic cell line of the pre-B phenotype, Int. J. Cancer 23 (1979) 174– 180. [9] A. Fossa, R. Siebert, H.C. Aasheim, G.M. Maelandsmo, A. Berner, S.D. Fossa, E. Paus, E.B. Smeland, G. Gaudernack, Identification of nucleolar protein No55 as a tumour-associated autoantigen in patients with prostate cancer, Br. J. Cancer 83 (2000) 743–749. [10] E. Munthe, E. Rian, T. Holien, A. Rasmussen, F.O. Levy, H. Aasheim, Ephrin-B2 is a candidate ligand for the Eph receptor, EphB6, FEBS Lett. 466 (2000) 169–174. [11] A. Simonsen, R. Lippe, S. Christoforidis, J.M. Gaullier, A. Brech, J. Callaghan, B.H. Toh, C. Murphy, M. Zerial, H. Stenmark, EEA1 links PI(3)K function to Rab5 regulation of endosome fusion, Nature 394 (1998) 494–498. [12] S.F. Altschul, T.L. Madden, A.A. Schaffer, J. Zhang, Z. Zhang, W. Miller, D.J. Lipman, Gapped BLAST and PSI-BLAST: a new generation of protein database search programs, Nucleic Acids Res. 25 (1997) 3389–3402. [13] Q.H. Zhang, M. Ye, X.Y. Wu, S.X. Ren, M. Zhao, C.J. Zhao, G. Fu, Y. Shen, H.Y. Fan, G. Lu, M. Zhong, X.R. Xu, Z.G. Han, J.W. Zhang, J. Tao, Q.H. Huang, J. Zhou, G.X. Hu, J. Gu, S.J. Chen, Z. Chen, Cloning and functional analysis of cDNAs with open reading frames for 300 previously undefined genes expressed in CD34+ hematopoietic stem/progenitor cells, Genome Res. 10 (2000) 1546–1560.
Characterization of the novel human transmembrane protein 9 (TMEM9) that localizes to lysosomes and late endosomesq Marit Kveine, Ellen Tenstad, Guri Døsen, Steinar Funderud, and Edith Rian* Department of Immunology, The Norwegian Radium Hospital, 0310 Oslo, Norway Received 26 August 2002
Abstract We have identified and characterized the novel human transmembrane protein 9 (TMEM9). TMEM9 encodes a 183 amino-acid protein that contains an N-terminal signal peptide, a single transmembrane region, three potential N-glycosylation sites, and three conserved cys-rich domains in the N-terminus, but no hitherto known functional domains. The protein is highly conserved between species from Caenorhabditis elegans to man and belongs to a novel family of transmembrane proteins. The TMEM9 gene consists of at least 6 exons and is localized to chromosome 1q41. TMEM9 mRNA is expressed in a wide range of tissues and cells. COS-1 cells transfected with a TMEM9 expression plasmid gave three bands of about 28, 31, and 33 kDa representing glycosylated forms of TMEM9 with a protein backbone of about 26 kDa. In COS-1 cells transfected with a TMEM9-GFP expression construct ,TMEM9GFP is co-expressed with LAMP1 on late endosomes and lysosomes as well as on ER. Thus, TMEM9 is a phylogenetically conserved, widely expressed transmembrane protein with a potential, but unknown function in intracellular transport. Ó 2002 Elsevier Science (USA). All rights reserved. Keywords: Transmembrane protein type I; Signal sequence trap; Novel protein family; Late endosomes; Lysosomes; Chromosome 1
B lymphopoiesis is a stepwise process consisting of initially, a stroma-dependent phase including adherence to and exchange of factors with the cells in the bone marrow microenvironment, and subsequently, a stromaindependent part where the cells further differentiate and migrate out of the bone marrow to peripheral lymphoid organs [1]. Our main focus is to identify and characterize proteins important for the human stroma-dependent B cell differentiation. To identify genes encoding secreted and membrane expressed proteins in a pro-B cell line, we therefore employed a signal sequence trap (SST) based on retroviral expression (SST-REX) [2]. SSTs are methods used to clone and identify the cDNAs of proteins that are guided to the secretory pathways and cellular membranes by the their N-terminal, hydrophobic signal peptide [3–5]. Thus, not only secretory and plasma membrane-bound factors, but also proteins on internal membranes (ER, Golgi network, mitochondria q
TMEM9 GenBank Accession Number is AY138587. * Corresponding author. Fax: +47-22-500-730. E-mail address: [email protected] (E. Rian).
and organelles of the endocytic pathway), can be identified by a signal sequence trap. However, the latter may also be important for cellular communication, in particular proteins on endosomes and lysosomes that may participate in endocytosis of ligand-bound receptors and receptor degradation or recycling [6]. We report here the cloning and identification by SSTREX, and the subsequent characterization of a novel human transmembrane protein, TMEM9, belonging to a new protein family. The mRNA is widely expressed, including in B cells and other hematopoietic cells. Transfected cells produce three glycosylated forms of TMEM9 and the protein co-localizes with LAMP1 on late endosomes and lysosomes, suggesting a role for TMEM9 in intracellular transport.
Materials and methods Reagents and antibodies. The following antibodies were used: mouse anti-human myc antibody (9E10 hybridoma, [7]), rabbit antimouse IgG-HRP (Dako AS, Glostrup, Denmark), mouse anti-human LAMP1 and EEA1 antibodies (provided by H. Stenmark, Department
0006-291X/02/$ - see front matter Ó 2002 Elsevier Science (USA). All rights reserved. PII: S 0 0 0 6 - 2 9 1 X ( 0 2 ) 0 2 2 2 8 - 3
M. Kveine et al. / Biochemical and Biophysical Research Communications 297 (2002) 912–917 of Biochemistry, The Norwegian Radium Hospital), and Cy3-conjugated goat anti-mouse and goat anti-rabbit antibodies (Amersham Pharmacia Biotech, Uppsala, Sweden). Tunicamycin, saponin, phenylmethylsulfonyl fluoride (PMSF), Tween 20, and aprotinin were supplied from Sigma Chemicals (Steinheim, Germany). Polybrene was purchased from Speciality Media (Phillipsburg, USA) and IL-3 was from R&D systems (Abingdon, GB). Cells. The following human cell lines were used in this study: the pro-B cell lines TOM-1 and BV173 (both provided by Dr. T. Logtenberg, Utrecht, The Netherlands), the pre-B cell lines Reh (ATCC CRL-8286) and Nalm 6 [8], the mature B cell lines Bjab (provided by G. Kline, Basel Institute for Immunology, Basel, Switzerland) and Daudi (ATCC CCL-213), the plasmacytoid cell lines U266 (ATCC TIB 196), the T cell lines Jurkat (ATCC TIB-152) and HPB-ALL (provided by M. Greaves, Leukaemia Research Fund Centre, Institute of Cancer Research, London, United Kingdom), the myeloid cell lines KG1-A (ATCC CCL-246), HL-60 (ATCC CCL-240), and U937 (ATCC CRL-1593), the erythroid precursor cell line K562 (ATCC CCL-243), the human larynx carcinoma HEp-2 (ATCC CCL 23) and COS-1 (ATCC CRL1650). All cell lines were grown in RPMI 1640 medium supplemented with 10% FCS, except COS-1 and HEp-2 cells that were grown in DulbeccosÕs modified EagleÕs medium (DMEM) supplemented with 10% FCS. All the cells were cultured at 37 °C in a humidified atmosphere with 5% CO2 . SST-REX cDNA library construction and signal sequence trap performance. Total RNA was prepared from the human pro-B lymphoid cell line BV173 by standard guanidine thiocyanate methods and polyðAÞþ RNA was further purified by oligo(dT)-beads (Dynal, Oslo, Norway). cDNA was synthesized from 5 lg polyðAÞþ RNA using random hexamers and the Stratagene cDNA synthesis kit (Stratagene, Cedar Creek, USA) according to manufacturerÕs instructions. BstXI adapters (Invitrogen, Carlsbad, CA) were ligated to the blunted cDNA ends before separation through SizeSep 400 Spin Column (Pharmacia, Uppsala, Sweden) and insertion of the cDNA into BstXI sites of the retroviral vector pMX-SST. The cDNA library was electroporated into competent XL1-blue bacteria. Plasmid DNA was prepared using the Quiagen Maxi kit (Qiagen, Hilden, Germany). The SST-REX was performed as described by Kitamura [2] with some modifications. Briefly, the BV173 cDNA library was converted to retrovirus by transfection into the packaging cell line Phoenix Eco (G.P. Nolan, Stanford University, CA, USA) using Fugene transfection agent (Roche, Basel, Switzerland). The supernatant contained 106 viral particles per ml as determined from the titer of the control vector pMX-GFP on NIH3T3 cells in a parallel experiment. 0:3 106 viral particles were used to infect 1:2 106 mycoplasma free Ba/F3 cells (DSMZ, Braunschweig, Germany) by spinning the cells (500g, 1 h) in the presence of Polybrene (4 lg/mL) and IL-3 (3 ng/mL). The resulting infection efficiency in Ba/F3 cells was 10% as assessed from the infection efficiency of a control vector pMX-GFP in Ba/F3 cells. The cells were deprived of IL-3 the following day and transferred to 96-well multititer plates to reveal the IL-3 independent clones. Integrated cDNAs from these Ba/F3 clones were PCR-amplified using vector primers on genomic DNA (QiaAmp DNA Blood Minikit, Qiagen, Hilden, Germany) and sequenced. Cloning of TMEM9 cDNA. One of the novel clones identified by SST-REX showed full homology to a DNA encoding a hypothetical protein clone MGC:891. Analysis of the ORF indicated a transmembrane type I protein and the clone was, therefore, chosen for further characterization. The putative protein coding part including 5 bp upstream of the predicted ATG was PCR amplified from a prostate cDNA library (provided by A. Foss a, [9]) using site specific primers (forward primer 50 -TAG GAA GCT TTA AGC ATG AAG CTC TTA-30 and reverse primer for vector with myc-His epitopes 50 -ATC CTC GAG CGG CCG CCG CTG AGC ATC TTG TGC CGA-30 and reverse primer for the pEGFP-N3 vector 50 -ATC CTC GAG GAT CCG CTG AGC ATC TTG TGC CGA-30 ). The primers included restriction enzyme sites HindIII, NotI, and BamHI, respectively.
913
The TMEM9 cDNA was either ligated in-frame upstream of the mycHis epitope cDNA in pcDNA3.1/myc-His (Invitrogen, Carlsbad, USA) or in-frame upstream of green fluorescent protein (GFP) in pEGFP-N3 (Clontech, Palo Alto, USA). The constructs were used to produce mycHis-tagged TMEM9 and TMEM9-GFP fusion protein, respectively. Northern blot analysis. Total RNA was isolated by the RNeasy mini kit (Qiagen, Hilden, Germany). Twenty lg total RNA from each sample was loaded in each well and subjected to Northern blot analysis as previously described [10]. The membranes were hybridized overnight with a 32 P-dCTP-labeled TMEM9 cDNA probe (550 bp ORF fragment, isolation described above), thereafter stripped and hybridized with a probe against 18S rRNA (GenBank Accession No. AJ271813) to monitor loading differences. Two human multiple tissues Northern blots (Clontech H2 and H3, Palo Alto, USA) were hybridized with the TMEM9 cDNA probe according to the instructions of the manufacturer. Signals were detected by phosphoimager (Molecular Dynamics, Amersham Pharmacia Biotech, UK). Western blot analysis. Plasmids encoding TMEM9 tagged with the myc-His epitopes or vector alone (pcDNA3.1/myc-His) were transfected into COS-1 cells using Lipofectamine (Invitrogen, Carlsbad, USA). One to two days post-transfection, 1 106 cells were lysed with 0.05% saponin in PBS supplemented with 30 lg/ml aprotinin and 1 mM phenylmethylsulfonyl fluoride. For tunicamycin-treatment, the transfected cells were incubated with tunicamycin (5 lg/ml) or vehicle (0.025% DMSO) for 16 h before cell lysis. Near equal amounts of total protein extract (related to cell number) were denatured (475 mM bmercaptoethanol) and separated on a 10% polyacrylamide gel, followed by immunoblotting using mouse anti-myc as primary antibody and rabbit anti-mouse HRP as secondary antibody. The proteins were visualized with the ECL system (Amersham, Uppsala, Sweden). The loading was evaluated by Ponceau red staining of the filters. Confocal fluorescence microscopy. HEp-2 cells were grown on coverslips and transiently transfected with TMEM9-GFP in pEGFP-N3 or pEGFP-N3 as control. The transfection reagent Fugene was used according to producerÕs protocol. One day post-transfection, the cells were fixed with 3% paraformaldehyde and subsequently incubated with antiLAMP1 antibody, followed by a secondary antibody (Cy3-labeled antimouse antibody) in a solution containing PBS and 0.05% saponin as previously described [11]. The cellular localization of TMEM9 was observed by examination of coverslips with a Leica TCS NT confocal scanning microscope system equipped with a Kr/Ar laser and a PL Fluotar 100/1.30 oil immersion objective (Leica Microsystems, Heidelberg, Germany). Appropriate emission filter settings and controls were included to exclude bleed-through effects.
Results and discussion Cloning and structural characterization To identify factors important for human stromadependent B lymphopoiesis, we have employed SST-REX [2] on human pro-B lymphoma cells BV173 as a model system as these cells adhere to bone marrow stroma cells and partly differentiate to IgM-positive cells in stroma coculture (not shown). RNA from this cell line was used to construct a cDNA library in the pMX-SST vector, and a preliminary screening employing the SST-REX method [2] resulted in the identification of 18 different known cDNAs and 4 which represented novel sequences. One of the novel sequences with a potential open reading frame (ORF) with properties of a transmembrane type I protein (see below) was selected for further characterization.
914
M. Kveine et al. / Biochemical and Biophysical Research Communications 297 (2002) 912–917
Fig. 1. (A) Amino-acid sequence alignment of TMEM9 from human (hTMEM9), mouse (mTMEM9), and Tokifugu rubripes (trTMEM9) with the human homologous hypothetical protein hC11ORF15 and its murine orthologous hypothetical protein mD7H11ORF15, as well as with the orthologous hypothetical proteins from C. elegans T16732 and D. melanogaster AAF52042. The alignment was obtained by the use of ClustalW (http:// www.ch.embnet.org/software/ClustalW.html). Black shading indicates identities, gray shading indicates similarities. Potential N-glycosylation sites are marked with dashes. The predicted signal peptide cleavage site is marked with an arrow, and the suggested transmembrane (TM) region is marked with a line. Cystein-rich domains are marked with asterisks. (B) Phylogenetic analysis of hTMEM9 and the related proteins aligned in (A). The dendrogram was obtained by using Protpars in the Phylip program package (http://bioweb.pasteur.fr/seqanal/phylogeny/phylip-uk.html). (C) The genomic structure characterized of at least 6 exons represented by gray boxes (not on scale). The translational start and stop codons are marked.
Homology search performed with the BLAST program [12] (http://www.ncbi.nlm. nih.gov) showed that the cDNA of 313 bp was identical to that of the hypothetical
protein, clone MGC:891 (GenBank Accession No. BC001106, Hs.181444), the uncharacterized DERP4 (Accession No. AB013909), and clone HSPC186
M. Kveine et al. / Biochemical and Biophysical Research Communications 297 (2002) 912–917
identified in a hematopoietic stem and progenitor cell cDNA library [13]. The cDNA included the ATG translation start codon and a 62 bp upstream sequence. Sequence information from clone MGC:891 was used to construct primers and the protein coding part of the cDNA was amplified from a prostate cDNA library. The cDNA encodes an ORF of 183 amino-acid (aa) residues (Fig. 1A). Analysis of the predicted protein sequence using the SignalP, PSORTII, and ScanProsite programs (http://www.expasy.ch) indicated a protein with an Nterminal signal peptide and a single transmembrane spanning region (transmembrane protein type I) (Fig. 1A). Three N-glycosylation sites in the N-terminal part of the deduced mature protein were suggested (dashes, Fig. 1A). The first site, however, is not likely to be glycosylated, since it is predicted to be the first amino acid in the mature protein. Apart from these features, no known protein domains or motifs pointing to a function or cellular localization of this protein were found. The protein was named transmembrane protein 9 (TMEM9, approved by the Human Genome Nomenclature Committee (http:// www.gene. ucl.ac.uk/nomenclature). The sequence data has GenBank entry AY138587. Database homology searches identified one hypothetical human homolog with 57% aa identity (74% positivity) to TMEM9 (C11orf15, Accession No. XP_011973) (Fig. 1A). Additionally, hypothetical orthologs in several species were located; mouse (mTMEM9, 95% aa identity, 96% positivity, Accession No. BAB23903), Takifugu rubripes (trTMEM9, 76% aa identity, 85% positivity, Accession No. T30536), Drosophila melanogaster (31% aa identity, 48% positivity, Accession No. AAF52042) and Caenorhabditis elegans (28% aa identity, 46% positivity, Accession No. T16732). Finally, the human homolog has a hypothetical mouse ortholog (D7H11orf15, Accession No. BAB25033) showing 56% aa identity (73% positivity) to TMEM9 (see later). We note that Zhang et al. [13] have also assigned the putative ORF of clone HSPC186 as a potential transmembrane protein with several orthologs. Alignment of the ORFs by ClustalW demonstrates several regions of very high homology (Fig. 1A). First, the suggested transmembrane region of TMEM9 (approximately aa 90–112) is highly homologous between all hypothetical proteins, indicating that all seven are transmembrane proteins. Similar transmembrane topology is suggested by PSORTII for all proteins, except for the C. elegans version (T16732), which may have three transmembrane spanning regions and an uncleavable N-terminus. Second, the C-terminal, cytoplasmic regions of the seven ORFs are nearly identical, suggesting that this domain is functionally important. Third, three conserved cystein-rich domains are found in the N-terminal regions (marked with asterisk, Fig. 1A). These regions may participate in protein folding, protein interactions, and multimerization. Finally, the suggested N-glycosylation sites are partly conserved, and thus, of
915
potential importance for the protein structure and function. Phylogenetic analysis of the protein sequences reflects the close evolutionary relationship between these proteins (Fig. 1B). The high degree of homology between species indicates that these proteins build a novel family of transmembrane proteins highly conserved throughout the phylogenesis, and with potentially important function(s) in all these organisms. Chromosomal localization and genomic organization BLAST with the TMEM9 cDNA against the human genome database (http://www.ncbi.nlm.nih.gov) shows that the TMEM9 gene is localized to chromosome 1q41 (LOC51235). Sequence information from Homo sapiens chromosome 1 clone RP5-894H24 (Accession No. AL 139159), currently undergoing sequencing, suggests that the gene spans at least a 9 kb sequence and that the genomic structure of the gene is characterized by at least 6 exons (Fig. 1B). The ATG translation start site is found on exon 2, whereas the translation stop codon and a potential ATTAAA polyadenylation signal is on exon 6. mRNA expression analysis Northern blot analysis of human tissue RNA blots shows that a major 1.7 kb transcript and a minor 1.1 kb form of TMEM9 mRNA are expressed in a wide range of tissues (Fig. 2A). The minor transcript may represent an alternatively spliced form. TMEM9 mRNA is abundant in adrenal gland, thyroid gland, ovary, testis, and prostate, and to a lesser degree expressed in trachea, spinal cord, stomach, colon, small intestine, thymus, and spleen (Fig. 2A). Low expression is observed in bone marrow, lymph node, and peripheral blood lymphocytes (PBL). However, a more thorough analysis of cell lines of hematopoietic origin demonstrates TMEM9 mRNA expression in cell lines of B lineage cell origin (Tom-1, BV173, Reh, Nalm6, Bjab, Daudi, and U266), T lineage cell origin (HPB-All, Jurkat, and JM), myeloid origin (KG1a, HL60, and U937), and erythroid origin (K562) (Fig. 2B). There is no obvious correlation between the differentiation level represented by the varying cell lines and the mRNA expression level of TMEM9. Western blot analysis and protein deglycosylation We verified that the TMEM9 ORF results in a translated product by producing TMEM9 with a C-terminal myc-His epitope tag (Fig. 3). Total cell lysates from transfected COS-1 cells were analyzed by Western blotting, demonstrating a band triplet of approximately 28, 31, and 33 kDa (arrows). The three bands most likely represent glycosylated forms of the TMEM9 protein, since several potential N-glycosylation sites are found in the amino-acid sequence. Additionally, a heavier band of
916
M. Kveine et al. / Biochemical and Biophysical Research Communications 297 (2002) 912–917
Fig. 2. Northern blot analysis of TMEM9 mRNA expression. (A) Commercial human tissue blots show expression of two TMEM9 transcripts of approximately 1.7 and 1.1 kb. Molecular weight markers are to the right. (B) Expression of TMEM9 mRNA in hematopoietic cell lines of B lineage cell origin (Tom-1, BV173, Reh, Nalm6, Bjab, Daudi, and U266), T lineage cell origin (HPB-All, Jurkat, and JM), myeloid (KG1a, HL60, and U937), and erythroid origin (K562). Probe against 18S rRNA reveals loading differences. The figure is composed of lanes from a single filter exposure.
approximately 60 kDa was routinely observed. We assume that this band may represent a dimer form of the protein formed by interaction between one or more of the
Fig. 3. Western blot analysis of recombinant, myc-His-tagged TMEM9. COS-1 cells were transfected with a plasmid containing the TMEM9 myc-His construct or empty vector (pcDNA3.1/myc-His). After transfection the cells were treated with tunicamycin (T) or vehicle (C) before cell lysis and Western blot analysis. The arrows point out the three glycosylated forms of TMEM9. Molecular weight markers are to the left.
cystein-rich domains found in the N-terminal part. To test whether the protein bands were glycosylated forms of TMEM9, we treated the transfected cells with tunicamycin (5 lg/ml), which inhibits N-glycosylation in the ER. Tunicamycin-treated cells produced a major, single 26 kDa form of myc-His tagged, TMEM9 as well as a weak band of approximately 52 kDa. The molecular masses correspond to the theoretical Mw of the deduced protein sequence of 20.6 kDa, with an addition of 3.7 kDa representing the tag and a dimer form of the tagged protein, respectively. The different forms of recombinant TMEM9 with the myc-His-tag were also found in HEp-2 cells, and both COS-1 and HEp-2 cells produced a similar pattern of glycosylated TMEM9-GFP fusion protein (data not shown). Thus, these data suggest that TMEM9 exists in three glycosylated forms that may have distinct functional and/or affinity properties. Parts of the glycosylation pattern may result from unprocessed carbohydrate chains that render in the ER due to overloading of the transfected system (see below). Cellular localization Since proteins with N-terminal signal peptides may localize to internal cell membranes, as well as the plasma membrane and to secretory vesicles, it is important to determine the cellular distribution of novel proteins identified by signal sequence trap. PSORTII analysis of the protein sequence gave no clue as to which cellular membrane the protein can be directed to. We, therefore, transfected COS-1 cells with a plasmid containing the TMEM9-GFP construct. The empty pEGFP-N3 plasmid and mock transfected cells were used as negative controls (not shown). Confocal microscopy of transfected cells demonstrated fluorescence staining in ER and cytoplasmic vesicles that co-stained with an antiLAMP1 antibody, which is a marker for late endosomes and lysosomes (Fig. 4). No GFP staining was apparent in the nucleus (Fig. 4). TMEM9-GFP did not co-localize with EEA1, a marker for early endosomes (not shown). This indicates that late endosomes and lysosomes are the cellular localization sites for TMEM9. We assume that the strong fluorescence observed in ER is a result of overproduction and accumulation of the protein in transiently transfected cells. Cells transfected with the control plasmid (pEGFP-N3) showed staining mainly in the nucleus, whereas mock transfected cells and cells treated with secondary antibody alone showed no fluorescence staining (not shown). In conclusion, we have characterized a novel human gene and gene product TMEM9. TMEM9 is a type I transmembrane protein that belongs to a novel and phylogenetically conserved family of transmembrane proteins. The gene, consisting of at least 6 exons, is localized to chromosome 1q41. TMEM9 mRNA is widely expressed and transfected cells produce three glycosylated
M. Kveine et al. / Biochemical and Biophysical Research Communications 297 (2002) 912–917
917
Fig. 4. Confocal fluorescence microscopy of fixed COS-1 cells transfected with the TMEM9-GFP expression plasmid. (A) A cell expressing TMEM9GFP shows staining in ER and vesicular structures. (B) Immunostaining of the same cell with anti-LAMP1 antibody. (C) Merged image where yellow staining shows co-localization of TMEM9-GFP and LAMP1.
forms of a myc-His-tagged TMEM9 protein. TMEM9GFP fusion protein partly co-localizes with LAMP1 on late endosomes and lysosomes. The function of TMEM9 remains to be elucidated, but the cellular localization suggests that TMEM9 may participate in intracellular transport.
Acknowledgments We thank H. Stenmark and C. Raiborg, Department of Biochemistry, The Norwegian Radium Hospital, Norway, for providing the LAMP1 and EEA1 antibodies and for invaluable help with the confocal laser scanning microscopy experiments. We are grateful to T. Kitamura, Department of Hematopoietic Factors, University of Tokyo, Japan, for sharing the pMX-SST plasmid and experience regarding the SST-REX procedure with us, and to G.P. Nolan, Stanford University, CA, USA, for the Phoenix Eco packaging cell line. Thanks to A. Foss a, Department of Immunology, The Norwegian Radium Hospital, for providing the DU145 prostate cDNA library and to E.B. Smeland, Department of Immunology, The Norwegian Radium Hospital, for carefully reading the manuscript.
References [1] T.W. LeBien, Fates of human B-cell precursors, Blood 96 (2000) 9–23. [2] T. Kojima, T. Kitamura, A signal sequence trap based on a constitutively active cytokine receptor, Nat. Biotechnol. 17 (1999) 487–490. [3] E. Rian, T. Wabakken, H. Hauge, H.C. Aasheim, A signal sequence trap based on cell enrichment using anti-CD19 antibody coated magnetic beads, Scand. J. Immunol. 54 (2001) 280–284.
[4] R.D. Klein, Q. Gu, A. Goddard, A. Rosenthal, Selection for genes encoding secreted proteins and receptors, Proc. Natl. Acad. Sci. USA 93 (1996) 7108–7113. [5] K. Tashiro, T. Nakamura, T. Honjo, The signal sequence trap method, Methods Enzymol. 303 (1999) 479–495. [6] S. Mukherjee, R.N. Ghosh, F.R. Maxfield, Endocytosis, Physiol. Rev. 77 (1997) 759–803. [7] G.I. Evan, G.K. Lewis, G. Ramsay, J.M. Bishop, Isolation of monoclonal antibodies specific for human c-myc proto-oncogene product, Mol. Cell Biol. 5 (1985) 3610–3616. [8] R. Hurwitz, J. Hozier, T. LeBien, J. Minowada, K. GajlPeczalska, I. Kubonishi, J. Kersey, Characterization of a leukemic cell line of the pre-B phenotype, Int. J. Cancer 23 (1979) 174– 180. [9] A. Fossa, R. Siebert, H.C. Aasheim, G.M. Maelandsmo, A. Berner, S.D. Fossa, E. Paus, E.B. Smeland, G. Gaudernack, Identification of nucleolar protein No55 as a tumour-associated autoantigen in patients with prostate cancer, Br. J. Cancer 83 (2000) 743–749. [10] E. Munthe, E. Rian, T. Holien, A. Rasmussen, F.O. Levy, H. Aasheim, Ephrin-B2 is a candidate ligand for the Eph receptor, EphB6, FEBS Lett. 466 (2000) 169–174. [11] A. Simonsen, R. Lippe, S. Christoforidis, J.M. Gaullier, A. Brech, J. Callaghan, B.H. Toh, C. Murphy, M. Zerial, H. Stenmark, EEA1 links PI(3)K function to Rab5 regulation of endosome fusion, Nature 394 (1998) 494–498. [12] S.F. Altschul, T.L. Madden, A.A. Schaffer, J. Zhang, Z. Zhang, W. Miller, D.J. Lipman, Gapped BLAST and PSI-BLAST: a new generation of protein database search programs, Nucleic Acids Res. 25 (1997) 3389–3402. [13] Q.H. Zhang, M. Ye, X.Y. Wu, S.X. Ren, M. Zhao, C.J. Zhao, G. Fu, Y. Shen, H.Y. Fan, G. Lu, M. Zhong, X.R. Xu, Z.G. Han, J.W. Zhang, J. Tao, Q.H. Huang, J. Zhou, G.X. Hu, J. Gu, S.J. Chen, Z. Chen, Cloning and functional analysis of cDNAs with open reading frames for 300 previously undefined genes expressed in CD34+ hematopoietic stem/progenitor cells, Genome Res. 10 (2000) 1546–1560.