The BTB-kelch Protein LZTR-1 Is a Novel Golgi Protein That Is Degraded upon Induction of Apoptosis
2005; Elsevier BV; Volume: 281; Issue: 8 Linguagem: Inglês
10.1074/jbc.m509073200
ISSN1083-351X
AutoresTanju G. Nacak, Kerstin Leptien, Doris Fellner, Hellmut G. Augustin, Jens Krøll,
Tópico(s)Lipid metabolism and biosynthesis
ResumoMembers of the BTB-kelch superfamily play important roles during fundamental cellular processes, such as the regulation of cell morphology, migration, and gene expression. The BTB-kelch protein LZTR-1 is deleted in the majority of DiGeorge syndrome patients and is believed to act as a transcriptional regulator. However, functional and expression profiling studies of LZTR-1 have not been performed thus far. Therefore, we examined the subcellular localization and function of LZTR-1 to gain insights into its biological role. Analysis of the primary structure of the protein revealed six N-terminal kelch motifs and two BTB/POZ domains at the C terminus within LZTR-1. Confocal analysis of the subcellular distribution of LZTR-1 using the Golgi markers GM130, Golgin-97, and TGN46 identified a localization of LZTR-1 exclusively on the cytoplasmic surface of the Golgi network that is mediated by its second BTB/POZ domain. In contrast to most other BTB-kelch proteins, LZTR-1 did not co-localize with actin. Treatment with brefeldin A did not lead to redistribution of LZTR-1 to the endoplasmic reticulum but caused its relocalization in dispersed, punctuated structures that were also positive for GM130. These data demonstrate that LZTR-1 is a Golgi matrix-associated protein. Upon induction of apoptosis, LZTR-1 was phosphorylated on tyrosine residues and subsequently degraded; that could be rescued partially by the addition of the caspase inhibitor Z-VAD-fmk and the proteasome inhibitors lactacystin and MG132. Taken together, our experiments identify LZTR-1 as the first BTB-kelch protein that exclusively localizes to the Golgi network, and the binding of LZTR-1 to the Golgi complex is mediated by its second BTB/POZ domain. Members of the BTB-kelch superfamily play important roles during fundamental cellular processes, such as the regulation of cell morphology, migration, and gene expression. The BTB-kelch protein LZTR-1 is deleted in the majority of DiGeorge syndrome patients and is believed to act as a transcriptional regulator. However, functional and expression profiling studies of LZTR-1 have not been performed thus far. Therefore, we examined the subcellular localization and function of LZTR-1 to gain insights into its biological role. Analysis of the primary structure of the protein revealed six N-terminal kelch motifs and two BTB/POZ domains at the C terminus within LZTR-1. Confocal analysis of the subcellular distribution of LZTR-1 using the Golgi markers GM130, Golgin-97, and TGN46 identified a localization of LZTR-1 exclusively on the cytoplasmic surface of the Golgi network that is mediated by its second BTB/POZ domain. In contrast to most other BTB-kelch proteins, LZTR-1 did not co-localize with actin. Treatment with brefeldin A did not lead to redistribution of LZTR-1 to the endoplasmic reticulum but caused its relocalization in dispersed, punctuated structures that were also positive for GM130. These data demonstrate that LZTR-1 is a Golgi matrix-associated protein. Upon induction of apoptosis, LZTR-1 was phosphorylated on tyrosine residues and subsequently degraded; that could be rescued partially by the addition of the caspase inhibitor Z-VAD-fmk and the proteasome inhibitors lactacystin and MG132. Taken together, our experiments identify LZTR-1 as the first BTB-kelch protein that exclusively localizes to the Golgi network, and the binding of LZTR-1 to the Golgi complex is mediated by its second BTB/POZ domain. The kelch motif was identified as a segment of 44–56 amino acids in the ORF1 protein in Drosophila (1Xue F. Cooley L. Cell. 1993; 72: 681-693Abstract Full Text PDF PubMed Scopus (375) Google Scholar). kelch proteins typically contain 4–7 kelch motifs that form a propeller-like structure. In addition to the kelch motifs, the majority of the kelch proteins contain a BTB/POZ domain (broad complex, tramtrack and brick a brac/Pox virus and zinc finger) that mediates protein-protein interactions (2Adams J. Kelso R. Cooley L. Trends Cell. Biol. 2000; 10: 17-24Abstract Full Text Full Text PDF PubMed Scopus (485) Google Scholar). Recently, the BACK domain (BTB and C-terminal kelch) has been described in several BTB-kelch proteins; however, its function is unknown (3Stogios P.J. Prive G.G. Trends Biochem. Sci. 2004; 29: 634-637Abstract Full Text Full Text PDF PubMed Scopus (132) Google Scholar). Bioinformatic analysis identified at least 71 kelch proteins in the human genome, which reveals that the kelch motif is widely distributed (4Prag S. Adams J.C. BMC Bioinformatics 2003. 2003; (4/42)Google Scholar). Many kelch domain-containing proteins interact with actin and are important mediators of fundamental cellular functions, such as regulation of cellular architecture, cellular organization, and cell migration (2Adams J. Kelso R. Cooley L. Trends Cell. Biol. 2000; 10: 17-24Abstract Full Text Full Text PDF PubMed Scopus (485) Google Scholar). In addition, the kelch motif-containing protein Keap1 can interact with the transcription factor Nrf2, which regulates the expression of downstream genes encoding detoxifying and antioxidant proteins (5Itoh K. Wakabayashi N. Katoh Y. Ishii T. Igarashi K. Engel J.D. Yamamoto M. Genes Dev. 1999; 13: 76-86Crossref PubMed Scopus (2792) Google Scholar, 6Zipper L.M. Mulcahy R.T. J. Biol. Chem. 2002; 277: 36544-36552Abstract Full Text Full Text PDF PubMed Scopus (290) Google Scholar). Recently, it has also been shown that some BTB-kelch proteins serve as a substrate-specific adapter for cullin 3 ubiquitin ligases (7Cullinan S.B. Gordan J.D. Jin J. Harper J.W. Diehl J.A. Mol. Cell. Biol. 2004; 24: 8477-8486Crossref PubMed Scopus (767) Google Scholar, 8Kobayashi A. Kang M.I. Okawa H. Ohtsuji M. Zenke Y. Chiba T. Igarashi K. Yamamoto M. Mol. Cell. Biol. 2004; 24: 7130-7139Crossref PubMed Scopus (1642) Google Scholar, 9Pintard L. Willis J.H. Willems A. Johnson J.L. Srayko M. Kurz T. Glaser S. Mains P.E. Tyers M. Bowerman B. Peter M. Nature. 2003; 425: 311-316Crossref PubMed Scopus (342) Google Scholar). Very little is known about the function of BTB-kelch proteins in vivo. So far, four BTB-kelch proteins have been knocked out in mice. Deletion of Keap1 results in postnatal lethality due to the constitutive activation of Nrf2 (10Wakabayashi N. Itoh K. Wakabayashi J. Motohashi H. Noda S. Takahashi S. Imakado S. Kotsuji T. Otsuka F. Roop D.R. Harada T. Engel J.D. Yamamoto M. Nat. Genet. 2003; 35: 238-245Crossref PubMed Scopus (692) Google Scholar). Male mice that are haploinsufficient for the BTB-kelch protein KLHL10 are infertile (11Yan W. Ma L. Burns K.H. Matzuk M.M. Proc. Natl. Acad. Sci. U. S. A. 2004; 101: 7793-7798Crossref PubMed Scopus (92) Google Scholar). Loss of the kelch domain-containing protein Nd1 leads to the lack of an overt phenotype in mice; however, it has been reported that Nd1 plays a protective role in doxorubicin-induced cardiotoxic responses (12Fujimura L. Matsudo Y. Kang M. Takamori Y. Tokuhisa T. Hatano M. Cardiovasc. Res. 2004; 64: 315-321Crossref PubMed Scopus (21) Google Scholar). Finally, we have recently shown that KLHL6 regulates B-lymphocyte receptor signaling and formation of germinal centers in mice (13Kroll J. Shi X. Caprioli A. Liu H.H. Waskow C. Lin K.M. Miyazaki T. Rodewald H.R. Sato T.N. Mol. Cell. Biol. 2005; 25: 8531-8540Crossref PubMed Scopus (56) Google Scholar). The BTB-kelch protein leucine zipper-like transcriptional regulator 1 (LZTR-1) was identified some years ago as a gene that is hemizygously deleted in the majority of DiGeorge syndrome patients (14Kurahashi H. Akagi K. Inazawa J. Ohta T. Niikawa N. Kayatani F. Sano T. Okada S. Nishisho I. Hum. Mol. Genet. 1995; 4: 541-549Crossref PubMed Scopus (56) Google Scholar). Symptoms of DiGeorge syndrome are congenital immunodeficiency and congenital heart disease, which are caused by a large deletion from chromosome 22 (15Lindsay E.A. Nat. Rev. Genet. 2001; 2: 858-868Crossref PubMed Scopus (215) Google Scholar, 16Scambler P.J. Hum. Mol. Genet. 2000; 9: 2421-2426Crossref PubMed Scopus (404) Google Scholar). Because of the weak homology of LZTR-1 to known members of the basic leucine zipper-like family, LZTR-1 has been proposed to act as a transcriptional regulator (14Kurahashi H. Akagi K. Inazawa J. Ohta T. Niikawa N. Kayatani F. Sano T. Okada S. Nishisho I. Hum. Mol. Genet. 1995; 4: 541-549Crossref PubMed Scopus (56) Google Scholar). However, the cellular localization and the biological function of LZTR-1 are not known. Therefore, we have pursued cellular localization and initial functional experiments to gain further insights into the role of LZTR-1. Here, we have characterized LZTR-1 on a cellular level and show that LZTR-1 is a novel Golgi matrix-associated protein. Furthermore, we show for the first time that the second BTB/POZ domain mediates the binding of LZTR-1 to the Golgi complex and that LZTR-1 is cleaved upon induction of apoptosis. Tissue Culture, Growth Factors, and Drugs—Human umbilical vein endothelial cells (HUVECs), 2The abbreviations used are: HUVEC, human umbilical vein endothelial cell; HMVEC, human dermal microvascular endothelial cell; HUASMC, human aortic smooth muscle cell; CHO, Chinese hamster ovary; RT, reverse transcriptase; TNFα, tumor necrosis factor-α; TRAIL, tumor necrosis factor-related apoptosis-inducing ligand; Z, benzyloxycarbonyl; fmk, fluoromethyl ketone; PARP, poly(ADP-ribose) polymerase; PBS, phosphate-buffered saline; GFP, green fluorescent protein; HEK, human embryonic kidney; ER, endoplasmic reticulum; bFGF, basic fibroblast growth factor; PDGF-BB, platelet-derived growth factor-BB; IPP, intracisternal A particle-promoted polypeptide; VEGF, vascular endothelial growth factor. human dermal microvascular endothelial cells (HMVEC), and human aortic smooth muscle cells (HUASMCs) were obtained from PromoCell (Heidelberg, Germany). HUVECs were grown in endothelial cell growth medium EGM-2 (PromoCell) supplemented with 10% fetal calf serum, and HMVEC were cultured in endothelial cell growth medium MV (PromoCell). HUASMCs were grown in medium no. 2 for smooth muscle cells (PromoCell). HeLa cells and HEK cells were grown in high glucose Dulbecco's modified Eagle's medium (Invitrogen) with 10% fetal calf serum and antibiotics. Chinese hamster ovary cells (CHO) were maintained in minimum essential α-medium (Invitrogen) with 10% fetal calf serum, and A375 melanoma cells were grown in Dulbecco's modified Eagle's medium (Invitrogen) with 10% fetal calf serum and antibiotics. Proteasome inhibitors MG132 and lactacystin were purchased from Merck Biosciences (Darmstadt, Germany) and caspase family inhibitor Z-VAD-fmk was obtained from Bio-Cat (Heidelberg, Germany). All inhibitors were dissolved in Me2SO and used at a final concentration of 50 μm. CoCl2 was purchased from Sigma and dissolved in PBS. VEGF, bFGF, and PDGF-BB were obtained from R&D Systems (Wiesbaden, Germany). Cloning of LZTR-1 and Generation of the Different Mutants—The full-length and the putative short isoform of LZTR-1 were cloned from MGC clone 21205 (Invitrogen) by PCR using Pfx polymerase (Invitrogen) and the following primers plus restriction sites: LZTR-1 full-length KpnI sense, 5′-CGG GGT ACC ATG GCT GGA CCG GGC AGC ACG-3′; LZTR-1 putative short isoform KpnI sense, 5′-CGG GGT ACC ATG GTG GCC TTT GAC CGC CAC-3′; and SacII antisense, 5′-TCC CCG CGG GAT GTC GGC GCC CAG CTC TGC-3′. The PCR products were cloned into the pGEM®-T vector (Promega, Mannheim, Germany) and subsequently released by KpnI and SacII digestion and ligated into the expression vector pcDNA3.1 MycHis (Invitrogen). Mutants were generated from the full-length LZTR-1 sequence using the following primers, subcloned into pGEM ®-T vector and subsequently ligated into the expression vector pcDNA3.1 MycHis. Primers for the mutants were as follows: mutant LZTR-1ΔBTB/POZ2 KpnI sense, 5′-GGT ACC ATG GCT GGA CCG GGC AGC-3′; ACG and SacII antisense, 5′-CCG CGG CAT GTC CTG GAT CAG AGA TGT GCC-3′; mutant LZTR-1Δkelch KpnI sense, 5′-GGT ACCATG CAG TTC TGC GAC GTG GAG TTC GTG-3′; SacII antisense, 5′-CCG CGG GAT GTC GGC GCC CAG CTC TGC-3′; mutant LZTR-1BTB/POZ2 KpnI sense, 5′-GGT ACC ATG CTG ATC CAG GAC ATG AAG GCA TAC CTG GAG-3′; SacII antisense, 5′-CCG CGG GAT GTC GGC GCC CAG CTC TGC-3′. The fusion protein GFP-BTB/POZ2, consisting of GFP and the 2.BTB/POZ domain of LZTR-1, was cloned by using the pCruz GFP™ expression vector (Santa Cruz Biotechnology, Heidelberg, Germany). The correct sequence of all constructs was confirmed by sequencing. Generation of a Polyclonal Antiserum—A polyclonal antiserum was generated against the peptide NH2-VQPSSDSEVGGAEVPE-CONH2, coupled to keyhole limpet hemocyanin and injected into rabbits (Eurogentec, Seraing, Belgium). Sera from final bleeds were affinity-purified against the peptide using the AminoLink ® Plus immobilization kit (Pierce). Immunostaining of Cells—Cells were seeded on coverslips and fixed in 100% methanol (-20 °C) for 5 min or 4% paraformaldehyde at room temperature for 30 min, respectively. Thereafter, paraformaldehyde-fixed cells were permeabilized for 10 min using 0.3% Triton X-100/PBS followed by a blocking step in 3% bovine serum albumin/PBS for 30 min for all cells. Cells were incubated with monoclonal anti-Myc antibody (clone 9E10), polyclonal rabbit anti-human LZTR-1, sheep anti-human TGN46 (Serotec, Duesseldorf, Germany), mouse anti-human Golgin-97 (Invitrogen), or mouse anti-human GM130 (BD Biosciences) antibodies for 90 min in 3% bovine serum albumin/PBS at room temperature. Cells were washed three times with PBS and then incubated for 1 h with goat anti-rabbit Alexa 488, goat anti-mouse Alexa 546, goat anti-sheep Alexa 546, or goat anti-mouse Alexa 488 (Invitrogen), respectively. Phalloidin was purchased from Invitrogen and incubated for 30 min on cells. Samples were analyzed with an Olympus IX50 fluorescence microscope (Olympus, Leinfelden-Echterdingen, Germany) or by confocal microscopy (LSM510 Axiovert 200M, Zeiss, Oberkochen, Germany). Northern Blot Analysis and RT-PCR—Total RNA from cell lines and mouse tissues was isolated using the RNeasy Kit (Qiagen, Hilden, Germany). The probe for the Northern blot (711 bp in length) was generated by PCR from the MGC clone using the following primers: sense 5′-ACC GCC ACC TCT ATG TGT GTT-3′ and antisense 5′-AGA GGA ACT GCA TGA GCA CCT-3′. PCR primers for mouse LZTR-1 generating a 307-bp product were: sense 5′-GGA CCC TTC GAA ACA GTG CAT-3′ and antisense 5′-ATG CTG CTC CCA TAG ACC ACA-3′. Transfection of HeLa and HEK Cells—Transient transfections of HeLa cells and HEK were performed in 6-well plates using Lipofectamine 2000 (Invitrogen). Cells (5 × 105) were seeded and transfected the next day using 6 μg of plasmid DNA and 4 μl of Lipofectamine 2000. Four hours later, the transfection medium was replaced by fresh growth medium, and cells were cultured for another 24 h. Induction of Apoptosis and Western Blot Analyses—HeLa cells were stimulated 1 day after LZTR-1 transfection with 2μm staurosporine (Roche Applied Science), 100 ng/ml TNFα (R&D Systems), or 100 ng/ml TRAIL (Chemicon, Hampshire, UK) in the presence of 2.5 μg/ml cycloheximide (Merck Biosciences) to induce apoptosis (17Bartke T. Pohl C. Pyrowolakis G. Jentsch S. Mol. Cell. 2004; 14: 801-811Abstract Full Text Full Text PDF PubMed Scopus (165) Google Scholar). Cells were lysed in radioimmune precipitation assay buffer and analyzed by Western blots using the anti-Myc antibody. To study induction of apoptosis by cytochrome c, LZTR-1-transfected HEK cells were lysed in Buffer A (20 mm HEPES, pH 7.5, 10 mm KCl,; 1.5 mm MgCl2, 1 mm EDTA, 1 mm dithiothreitol) and incubated in the presence of 10 μm cytochrome c from horse heart (Sigma) and 1 mm dATP for 1 h at 30 °C (17Bartke T. Pohl C. Pyrowolakis G. Jentsch S. Mol. Cell. 2004; 14: 801-811Abstract Full Text Full Text PDF PubMed Scopus (165) Google Scholar, 18Deveraux Q.L. Roy N. Stennicke H.R. Van Arsdale T. Zhou Q. Srinivasula S.M. Alnemri E.S. Salvesen G.S. Reed J.C. EMBO J. 1998; 17: 2215-2223Crossref PubMed Scopus (1242) Google Scholar), followed by anti-Myc Western blot. For immunoprecipitations, cell lysates were incubated for 120 min at 4 °C in presence of the anti-Myc antibody and protein G-agarose (Roche Applied Science) followed by Western blot using the anti-phosphotyrosine-specific antibody PY99 (Santa Cruz Biotechnology). Induction of apoptosis in HeLa cells and HEK cells was confirmed by Western blots using the PARP antibody (Biomol, Hamburg, Germany). The anti-actin antibody was purchased from Santa Cruz Biotechnology. Identification of LZTR-1—We have recently shown that BTB-kelch protein KLHL6 is specifically expressed in embryonic endothelial cells but not in adult endothelial cells in mice (13Kroll J. Shi X. Caprioli A. Liu H.H. Waskow C. Lin K.M. Miyazaki T. Rodewald H.R. Sato T.N. Mol. Cell. Biol. 2005; 25: 8531-8540Crossref PubMed Scopus (56) Google Scholar). To identify novel kelch proteins in the vascular system, we analyzed the expression of all 71 human kelch proteins (4Prag S. Adams J.C. BMC Bioinformatics 2003. 2003; (4/42)Google Scholar) by RT-PCR in different vascular cell types including endothelial cells and smooth muscle cells. As a result of this screen, we identified the expression of LZTR-1 in these cell types. LZTR-1 Is an Unusual BTB-kelch Protein—Human LZTR-1 consists of 840 amino acids and contains six kelch motifs and two BTB domains (Fig. 1). In contrast to all other known BTB-kelch-motif containing proteins, LZTR-1 has its kelch motifs in the N terminus and its BTB/POZ domains in the C terminus (Fig. 1). Furthermore, LZTR-1 lacks the BACK domain which is present in most other BTB-kelch proteins. Lastly, LZTR-1 contains an additional translational start site, as indicated by the Kozak sequences at position 289 (Fig. 1), suggesting that LZTR-1 may be expressed in two isoforms. Cloning and Expression of LZTR-1—We cloned full-length LZTR-1 as well as the putative short isoform of LZTR-1 (for details see "Experimental Procedures") and expressed both constructs in HeLa cells. Subsequently, we analyzed its expression by anti-Myc Western blot (Fig. 2A). The construct encoding the putative short isoform of LZTR-1 generates a signal with an apparent molecular mass of 65 kDa (Fig. 2A). In contrast, the construct encoding for the full-length sequence of LZTR-1 runs at 95 kDa but lacks an additional band at 65 kDa, which would result from the second translation start site within the fifth kelch motif at position 289 (Fig. 2A). These data indicate that the putative short isoform of LZTR-1 is not expressed. However, we cannot exclude the possibility that this second translation start site is used under other conditions. Next, we analyzed several mouse tissues (Fig. 2B) for the expression of LZTR-1 by RT-PCR. LZTR-1 was detectable in all tissues and organs indicating its ubiquitous expression. Moreover, we analyzed different human vascular cells and tumor cells for the expression of LZTR-1 and found LZTR-1 expression in all analyzed cell lines (Fig. 2C). Additionally, we stimulated the vascular cells with certain activators such as VEGF, bFGF, and PDGF-BB and with the hypoxia-mimicking agent CoCl2, but we did not observe differences in the intensity of expression (Fig. 2C). Taken together, the data indicate that LZTR-1 is a ubiquitously expressed molecule. Generation and Characterization of a LZTR-1-specific Antiserum— To pursue the cellular characterization of LZTR-1, we generated a polyclonal antiserum. We have chosen a peptide that is a part of the fifth kelch motif within LZTR-1 (Figs. 1 and 3A). To test the specificity and sensitivity of the antiserum, CHO cells were transfected with a C-terminal Myc-tagged full-length LZTR-1 construct, and anti-LZTR-1 and anti-Myc staining was performed. As shown in Fig. 3B, anti-Myc staining of LZTR-1-transfected CHO cells identified several positive cells, indicating the presence of the Myc-tagged LZTR-1 expression construct. Next, we used the affinity-purified LZTR-1 antiserum to co-stain LZTR-1 transfected CHO cells (Fig. 3B). The polyclonal antiserum recognized only the LZTR-1-transfected CHO cells and not the untransfected cells, as indicated by merged images of the anti-Myc and anti-LZTR-1 staining (Fig. 3B). To further demonstrate the specificity of the antiserum, we expressed the C-terminal Myc-tagged mutant LZTR1–1Δkelch, which lacks the antibody recognition site within LZTR-1. Expression of mutant LZTR-1Δkelch in CHO cells and subsequent staining using the anti-Myc and anti-LZTR-1 antibodies resulted in Myc-positive, LZTR-1-negative cells (Fig. 3B). These data demonstrate the specificity and sensitivity of the LZTR-1 antiserum. We also tested the ability of the antiserum to recognize LZTR-1 by Western blot analysis. Myc-tagged LZTR-1-transfected CHO cells, mock-transfected CHO cells, and CHO cells transfected with a Myc-tagged deletion mutant of angiopoietin-2 (19Maisonpierre P.C. Suri C. Jones P.F. Bartunkova S. Wiegand S.J. Radziejewski C. Compton D. McClain J. Aldrich T.H. Papadopoulos N. Daly T.J. Davis S. Sato T.N. Yancopoulos G.D. Science. 1997; 277: 55-60Crossref PubMed Scopus (2954) Google Scholar) were analyzed by Western blot using an anti-Myc and the anti-LZTR-1 antiserum. The anti-Myc antibody recognizes a 95-kDa signal in the LZTR-1-transfected CHO cells and a 30-kDa signal for angiopoietin-2 in transfected CHO cells (Fig. 3C, left panel). In contrast, the LZTR-1 antiserum only recognizes the 95-kDa band of LZTR-1 but not of angiopoietin-2 (Fig. 3C, right panel). These data show the specificity of the LZTR-1 antiserum in Western blot analysis. LZTR-1 Is a Novel Golgi Matrix-associated Protein—To analyze the cellular localization of LZTR-1, we stained several cell lines using the LZTR-1-specific antiserum. Initial results showed the localization of LZTR-1 in the perinuclear region of the cells, which prompted us to hypothesize that LZTR-1 localizes to the Golgi complex or to the endoplasmic reticulum (ER). To test this hypothesis, we examined LZTR-1 localization in HUASMC, HUVECs, and HeLa cells (Fig. 4A) by confocal microscopy. The LZTR-1 antiserum recognized distinct perinuclear structures that were also recognized by the anti-GM130 antibody (20Nakamura N. Rabouille C. Watson R. Nilsson T. Hui N. Slusarewicz P. Kreis T.E. Warren G. J. Cell Biol. 1995; 131: 1715-1726Crossref PubMed Scopus (676) Google Scholar), which marked the Golgi complex in all three cell types. The merged picture of antibody stainings clearly indicated co-localization of LZTR-1 with GM130 (Fig. 4A). Similar results were obtained by using anti-Golgin-97 antibody (21Barr F.A. Curr. Biol. 1999; 9: 381-384Abstract Full Text Full Text PDF PubMed Scopus (132) Google Scholar, 22Griffith K.J. Chan E.K. Lung C.C. Hamel J.C. Guo X. Miyachi K. Fritzler M.J. Arthritis Rheum. 1997; 40: 1693-1702Crossref PubMed Scopus (145) Google Scholar) (supplemental Fig. 1) and the Golgi marker TGN46 (23Prescott A.R. Lucocq J.M. James J. Lister J.M. Ponnambalam S. Eur J. Cell Biol. 1997; 72: 238-246PubMed Google Scholar) (data not shown) together with LZTR-1. In contrast, we did not observe a co-localization of LZTR-1 with the ER marker calreticulin (data not shown). Thus, LZTR-1 localizes to the Golgi network and is hereby identified as a new Golgi protein. In addition, LZTR-1 is the first known BTB-kelch protein that localizes exclusively to the Golgi complex. In contrast to many known BTB-kelch proteins (2Adams J. Kelso R. Cooley L. Trends Cell. Biol. 2000; 10: 17-24Abstract Full Text Full Text PDF PubMed Scopus (485) Google Scholar), LZTR-1 does not co-localize with actin, as examined by anti-LZTR-1 and phalloidin double staining and confocal microscopy (Fig. 4B), suggesting that it does not directly interact with actin. Next, we treated the cells for 60 min with brefeldin A, which blocks protein transport from the ER to the Golgi apparatus and induces morphological changes such as the collapse of the Golgi stacks. Moreover, in brefeldin A-treated cells, Golgi enzymes are redistributed back to the ER, most Coat proteins to the cytoplasm, and a growing number of Golgi matrix proteins to Golgi "remnants" and/or ER exit sites (24Hicks S.W. Machamer C.E. J. Biol. Chem. 2002; 277: 35833-35839Abstract Full Text Full Text PDF PubMed Scopus (66) Google Scholar). Subsequent analysis of the brefeldin A-treated cells showed a collapse and punctuated dispersion of the Golgi. Double staining using Golgi matrix protein marker GM130 together with LZTR-1 shows a complete merge of punctuate structures after brefeldin A treatment (Fig. 4A). Therefore, LZTR-1 does not redistribute to the ER but remains with the Golgi remnants. These data demonstrate that LZTR-1 localizes to the Golgi complex and identifies LZTR-1 as a Golgi matrix-associated protein. The Second BTB/POZ Domain Is Essential and Sufficient for the Localization of LZTR-1 to the Golgi—To identify the LZTR-1 region mediating the localization to the Golgi, we generated three deletion mutants of LZTR-1 (Fig. 5A). Mutant LZTR-1ΔBTB/POZ2 lacks the second BTB/POZ domain, mutant LZTR-1Δkelch lacks all six kelch motifs, and mutant LZTR-1BTB/POZ2 contains only the second BTB/POZ domain. We expressed these mutant proteins as well as full-length LZTR-1 in HeLa cells and identified their expression by an anti-Myc Western blot (Fig. 5B), demonstrating the correct expression of all transfected constructs. Next, we expressed the mutants and the full-length LZTR-1 clone in HeLa cells and subsequently analyzed the localization of the mutants by anti-TGN46 staining for Golgi localization and anti-Myc staining by confocal microscopy. Full-length LZTR-1 co-localized with TGN46 in HeLa cells (Fig. 5C). Mutant LZTR-1Δkelch and LZTR-1BTB/POZ2 showed the same staining pattern as the full-length LZTR-1. In contrast, mutant LZTR-1ΔBTB/POZ2 did not co-localize with the Golgi marker TGN46 and was widely distributed throughout the cytoplasm of the cells (Fig. 5C). To further prove binding of the second BTB/POZ domain of LZTR-1 to the Golgi complex, we generated a fusion protein consisting of GFP and the second BTB/POZ domain of LZTR-1. Overexpression in HeLa cells localized the GFP-BTB/POZ2 construct to the Golgi complex as indicated by GM130 staining (Fig. 5D, upper panel). In contrast, GFP alone resides in the cytoplasm and nucleus (Fig. 5D, lower panel). Therefore, the second BTB/POZ domain of LZTR-1 is necessary and sufficient for the localization of LZTR-1 to the Golgi complex. To the best of our knowledge, this is the first report showing that a BTB/POZ domain mediates a protein localization to the Golgi complex. Induction of Apoptosis by Staurosporine, TNFα, and TRAIL Induces Degradation of LZTR-1 by Caspase- and Proteasome-dependent Pathways—Recent studies have shown that the induction of apoptosis leads to the Golgi fragmentation that is characterized by the conversion of the Golgi ribbon to tubulovesicular membrane clusters (25Mancini M. Machamer C.E. Roy S. Nicholson D.W. Thornberry N.A. Casciola-Rosen L.A. Rosen A. J. Cell Biol. 2000; 149: 603-612Crossref PubMed Scopus (325) Google Scholar). Fragmentation of the Golgi complex is accompanied by the cleavage of some Golgi matrix proteins such as GRASP65, which has been identified as a substrate for caspase-3 (26Lane J.D. Lucocq J. Pryde J. Barr F.A. Woodman P.G. Allan V.J. Lowe M. J. Cell Biol. 2002; 156: 495-509Crossref PubMed Scopus (182) Google Scholar). To address whether LZTR-1 is cleaved in response to apoptosis induction, we incubated LZTR-1-transfected HeLa cells for 4 h with staurosporine, for 8 h with TNFα, or for 4 h with TRAIL plus cycloheximide to induce apoptosis (17Bartke T. Pohl C. Pyrowolakis G. Jentsch S. Mol. Cell. 2004; 14: 801-811Abstract Full Text Full Text PDF PubMed Scopus (165) Google Scholar). We then examined the status of the LZTR-1 protein by Western blot using the anti-Myc antibody. Treating cells with staurosporine, TNFα, or TRAIL plus cycloheximide induced degradation of LZTR-1 (Fig. 6A, upper panel). In contrast, incubating lysates of LZTR-1-transfected HEK cells with cytochrome c did not induce degradation of LZTR-1, indicating that degradation of LZTR-1 does not occur downstream of the mitochondrial pathway (Fig. 6A, upper panel). To further evaluate the degradation of LZTR-1, we analyzed different time points of staurosporine treatment and determined that 50% of LZTR-1 is cleaved after ∼60 min (Fig. 6B). In all of the degradation experiments, induction of apoptosis was confirmed by analyzing the lysates with an antibody recognizing the intact 116-kDa fragment (corresponding to nonapoptotic cells) and cleaved 85-kDa fragment (corresponding to apoptotic cells) of PARP (Fig. 6, A–D, middle panels). To further analyze the pathways that mediate the degradation of LZTR-1 upon apoptosis induction, LZTR-1-expressing HeLa cells were incubated for 1 h prior to stimulation with the caspase inhibitor Z-VAD-fmk (27Lowe M. Lane J.D. Woodman P.G. Allan V.J. J. Cell Sci. 2004; 117: 1139-1150Crossref PubMed Scopus (73) Google Scholar) and one of two proteasome inhibitors, MG132 or lactacystin (28Keller J.N. Markesbery W.R. J. Neurosci. Res. 2000; 61: 436-442Crossref PubMed Scopus (51) Google Scholar). In the presence of the caspase inhibitor Z-VAD-fmk, staurosporine-induced degradation of LZTR-1 was partially rescued (Fig. 6C). Furthermore, preincubation of LZTR-1-overexpressing HeLa cells with the proteasome inhibitors lactacystin and MG132 rescued the degradation of LZTR-1 upon staurosporine stimulation (Fig. 6C). These data show that LZTR-1 is cleaved by caspase- and proteasome-dependent pathways upon apoptosis induction. In addition, we determined whether LZTR-1 is phosphorylated upon induction of apoptosis, which represents a common pathway to unmask a normally hidden degradation signal within a protein. LZTR-1-transfected HeLa cells were stimulated with staurosporine for different amounts of time, and phosphorylation of LZTR-1 was determined by anti-phosphotyrosine Western blot. As shown in Fig. 6D, staurosporine induced a tyrosine phosphorylation of LZTR-1 after 30 min. It is interesting that phosphorylated LZTR-1 was not longer detectable after 60 min, although some LZTR-1 protein was still present (Fig. 6B). This shows that degradation of LZTR-1 is accompanied by its phosphorylation and suggests that tyrosine phosphorylation of LZTR-1 is a prerequisite for its degradation. Herein we have characterized the BTB-kelch protein LZTR-1 on the cellular level and report that: 1) LZTR-1 is ubiquitously expressed; 2) LZTR-1 is a novel Golgi matrix-associated molecule; 3) LZTR-1 is the first known BTB-kelch protein that exclusively localizes to the Golgi complex; 4) the second BTB/POZ domain of LZTR-1 mediates localization of LZTR-1 to the Golgi complex, therefore showing for the first time that a BTB/POZ domain is involved in protein localization to the Golgi complex; and 5) LZTR-1 is cleaved by caspases- and proteasome-dependent pathways upon induction of apoptosis. Members of the kelch protein superfamily show a high structural as well as functional diversity. The number of kelch motifs varies between four and seven, and the kelch motifs can be localized at the N and/or C terminus. Although kelch proteins have diverse activities, some members, mostly BTB-kelch proteins, interact with actin and play a role in actin stabilization and organization. In addition, other members of the kelch protein family show a cytoplasmic or nuclear localization and elicit different biological functions (2Adams J. Kelso R. Cooley L. Trends Cell. Biol. 2000; 10: 17-24Abstract Full Text Full Text PDF PubMed Scopus (485) Google Scholar). Therefore, a direct correlation between structure and function does not exist within this family. However, most known kelch proteins interact with structural components of the cell and regulate cellular architecture and cellular organization (2Adams J. Kelso R. Cooley L. Trends Cell. Biol. 2000; 10: 17-24Abstract Full Text Full Text PDF PubMed Scopus (485) Google Scholar). Here, we report the ubiquitous expression of the unusual BTB-kelch protein LZTR-1, which, in contrast to all known BTB-kelch proteins, has its six kelch motifs at the N terminus and its two BTB/POZ domains at the C terminus. Furthermore, LZTR-1 does not co-localize with actin, which indicates that it is not involved in the organization and stabilization of the cytoskeleton as it has been reported on other BTB-kelch proteins (2Adams J. Kelso R. Cooley L. Trends Cell. Biol. 2000; 10: 17-24Abstract Full Text Full Text PDF PubMed Scopus (485) Google Scholar). We examined the cellular localization of LZTR-1 and determined that this novel BTB-kelch protein co-localizes with the Golgi marker GM130, Golgin-97, and TGN46, indicating that LZTR-1 localizes to the Golgi complex. This is a clear difference from all other BTB-kelch proteins, which show mostly a cytoplasmic distribution (2Adams J. Kelso R. Cooley L. Trends Cell. Biol. 2000; 10: 17-24Abstract Full Text Full Text PDF PubMed Scopus (485) Google Scholar). Moreover, LZTR-1 is a Golgi matrix-associated protein because it was punctuated throughout the cytoplasm upon brefeldin A treatment and still co-localized with GM130. Taken together, the data identify LZTR-1 as the first known BTB-kelch protein that exclusively localizes to the Golgi complex. Thus far, only a portion of BTB-kelch protein Gigaxonin has been identified at the Golgi complex and ER (29Cullen V.C. Brownlees J. Banner S. Anderton B.H. Leigh P.N. Shaw C.E. Miller C.C. Neuroreport. 2004; 15: 873-876Crossref PubMed Scopus (14) Google Scholar). As a consequence, LZTR-1 could stabilize the Golgi complex as other BTB-kelch proteins stabilize cellular architecture by interacting with structural proteins (2Adams J. Kelso R. Cooley L. Trends Cell. Biol. 2000; 10: 17-24Abstract Full Text Full Text PDF PubMed Scopus (485) Google Scholar). In addition to describing the localization of LZTR-1 to the Golgi complex, we have also demonstrated that the second BTB/POZ domain of LZTR-1 mediates the localization to the Golgi. This is an unexpected finding, as for the first time it shows that a BTB/POZ domain mediates a Golgi complex localization. Furthermore, the data indicate that LZTR-1 is localized on the cytoplasmic surface of the Golgi complex because the truncated mutant LZTR-1ΔBTB/POZ2 was dispersed throughout the cytoplasm. Thus, we are establishing the BTB/POZ domain as a novel Golgi localization domain. So far, the BTB/POZ domain has been described as mediating a homomeric dimerization of proteins, and BTB/POZ-containing proteins act as transcriptional regulators (30Kelly K.F. Otchere A.A. Graham M. Daniel J.M. J. Cell Sci. 2004; 117: 6143-6152Crossref PubMed Scopus (44) Google Scholar). Moreover, it has been shown that the BTB/POZ domain defines a recognition motif for the assembly of substrate-specific RING/cullin/BTB ubiquitin ligase complex (31Geyer R. Wee S. Anderson S. Yates J. Wolf D.A. Mol. Cell. 2003; 12: 783-790Abstract Full Text Full Text PDF PubMed Scopus (260) Google Scholar). Morphological studies have shown that induction of apoptosis leads to the fragmentation of the Golgi complex (25Mancini M. Machamer C.E. Roy S. Nicholson D.W. Thornberry N.A. Casciola-Rosen L.A. Rosen A. J. Cell Biol. 2000; 149: 603-612Crossref PubMed Scopus (325) Google Scholar, 32Sesso A. Fujiwara D.T. Jaeger M. Jaeger R. Li T.C. Monteiro M.M. Correa H. Ferreira M.A. Schumacher R.I. Belisario J. Kachar B. Chen E.J. Tissue Cell. 1999; 31: 357-371Crossref PubMed Scopus (55) Google Scholar, 33Philpott K.L. McCarthy M.J. Becker D. Gatchalian C. Rubin L.L. Eur. J. Neurosci. 1996; 8: 1906-1915Crossref PubMed Scopus (22) Google Scholar). Although it is not clear why the Golgi complex is fragmented during apoptosis, a recent report shows that stacking protein GRASP65, but not GRASP55, is cleaved upon apoptosis induction and that cleavage of GRASP65 is required for Golgi fragmentation (26Lane J.D. Lucocq J. Pryde J. Barr F.A. Woodman P.G. Allan V.J. Lowe M. J. Cell Biol. 2002; 156: 495-509Crossref PubMed Scopus (182) Google Scholar). Furthermore, it has been suggested that other Golgi molecules are cleaved during apoptosis (26Lane J.D. Lucocq J. Pryde J. Barr F.A. Woodman P.G. Allan V.J. Lowe M. J. Cell Biol. 2002; 156: 495-509Crossref PubMed Scopus (182) Google Scholar). We have now identified a novel Golgi protein, LZTR-1, that is rapidly cleaved upon apoptosis induction by caspases- and proteasome-dependent pathways. In addition, LZTR-1 is phosphorylated on tyrosine residues, suggesting that phosphorylation of LZTR-1 provides a signal for its degradation. According to the idea that LZTR-1 is an important structural component of the Golgi complex, one could speculate that the cleavage of LZTR-1 is a requirement for the fragmentation of the Golgi complex. However, further experiments need to be done to prove this hypothesis. Interestingly, the BTB-kelch protein IPP (34Kim I.F. Mohammadi E. Huang R.C. Gene. 1999; 228: 73-83Crossref PubMed Scopus (38) Google Scholar) is not degraded upon apoptosis induction, 3T. G. Nacak and J. Kroll, unpublished data. indicating that degradation of BTB-kelch proteins upon apoptosis induction is not a common mechanism and suggesting a distinct role of LZTR-1 within the BTB-kelch protein superfamily during apoptosis. LZTR-1 was first described some years ago as a gene deleted in seven of eight DiGeorge syndrome patients. Furthermore, because of the weak homology of LZTR-1 to known members of the basic leucine zipper-like family, it has been proposed that LZTR-1 acts as a transcriptional regulator (14Kurahashi H. Akagi K. Inazawa J. Ohta T. Niikawa N. Kayatani F. Sano T. Okada S. Nishisho I. Hum. Mol. Genet. 1995; 4: 541-549Crossref PubMed Scopus (56) Google Scholar). However, because of its exclusive localization to the Golgi complex it seems unlikely that LZTR-1 acts as a transcriptional regulator. Taken together, we have identified LZTR-1 as the first BTB-kelch protein exclusively localized to the Golgi complex. Further studies will now be undertaken to reveal the exact function of LZTR-1 for the Golgi complex. We are grateful to Dr. Christopher L. Antos for the editing of this manuscript, to Dr. Ulrike Fiedler for providing the N-terminal Myc-tagged angiopoietin-2 mutant containing the fibrinogen-like domain of angiopoietin-2, and to Dr. Ru Chih C. Huang for providing the IPP plasmid. We are grateful to Christine Fulda for technical assistance. Download .pdf (.51 MB) Help with pdf files
Referência(s)