4.1R Proteins Associate with Interphase Microtubules in Human T Cells
2001; Elsevier BV; Volume: 276; Issue: 48 Linguagem: Inglês
10.1074/jbc.m107369200
ISSN1083-351X
AutoresCarmen M. Pérez-Ferreiro, Carlos M. Luque, Isabel Correas,
Tópico(s)Blood groups and transfusion
ResumoRed blood cell protein 4.1 (4.1R) is an 80-kDa protein that stabilizes the spectrin-actin network and anchors it to the plasma membrane. To contribute to the characterization of functional roles and partners of specific nonerythroid 4.1R isoforms, we analyzed 4.1R in human T cells and found that endogenous 4.1R was distributed to the microtubule network. Transfection experiments of T cell 4.1R cDNAs in conjunction with confocal microscopy analysis revealed the colocalization of exogenous 4.1R isoforms with the tubulin skeleton. Biochemical analyses using Taxol®(paclitaxel)-polymerized microtubules from stably transfected T cells confirmed the association of the exogenous 4.1R proteins with microtubules. Consistent with this, endogenous 4.1R immunoreactive proteins were also detected in the microtubule-containing fraction.In vitro binding assays using glutathioneS-transferase-4.1R fusion proteins showed that a constitutive domain of the 4.1R molecule, one that is therefore present in all 4.1R isoforms, is responsible for the association with tubulin. A 22-amino acid sequence comprised in this domain and containing heptad repeats of leucine residues was essential for tubulin binding. Furthermore, ectopic expression of 4.1R in COS-7 cells provoked microtubule disorganization. Our results suggest an involvement of 4.1R in interphase microtubule architecture and support the hypothesis that some 4.1R functional activities are cell type-regulated. Red blood cell protein 4.1 (4.1R) is an 80-kDa protein that stabilizes the spectrin-actin network and anchors it to the plasma membrane. To contribute to the characterization of functional roles and partners of specific nonerythroid 4.1R isoforms, we analyzed 4.1R in human T cells and found that endogenous 4.1R was distributed to the microtubule network. Transfection experiments of T cell 4.1R cDNAs in conjunction with confocal microscopy analysis revealed the colocalization of exogenous 4.1R isoforms with the tubulin skeleton. Biochemical analyses using Taxol®(paclitaxel)-polymerized microtubules from stably transfected T cells confirmed the association of the exogenous 4.1R proteins with microtubules. Consistent with this, endogenous 4.1R immunoreactive proteins were also detected in the microtubule-containing fraction.In vitro binding assays using glutathioneS-transferase-4.1R fusion proteins showed that a constitutive domain of the 4.1R molecule, one that is therefore present in all 4.1R isoforms, is responsible for the association with tubulin. A 22-amino acid sequence comprised in this domain and containing heptad repeats of leucine residues was essential for tubulin binding. Furthermore, ectopic expression of 4.1R in COS-7 cells provoked microtubule disorganization. Our results suggest an involvement of 4.1R in interphase microtubule architecture and support the hypothesis that some 4.1R functional activities are cell type-regulated. glutathioneS-transferase ezrin-radixin-moesin 1,4-piperazinediethanesulfonic acid 4-morpholineethanesulfonic acid Red blood cell protein 4.1, 4.1R or 4.1R80, was identified as an 80-kDa multifunctional protein of the membrane skeleton of human erythrocytes. In these cells, protein 4.1R stabilizes the spectrin-actin network and anchors it to the overlying lipid bilayer through interactions with cytoplasmic domains of transmembrane proteins (reviewed in Ref. 1Conboy J.G. Semin. Hematol. 1993; 30: 58-73PubMed Google Scholar). The formation of the spectrin-actin-4.1R ternary complex is essential for the maintenance of normal erythrocyte morphology and membrane mechanical strength, as alterations in the spectrin-actin binding site of 4.1R, located at the COOH-terminal region of the molecule (2Correas I. Speicher D.W. Marchesi V.T. J. Biol. Chem. 1986; 261: 13362-13366Abstract Full Text PDF PubMed Google Scholar, 3Correas I. Leto T.L. Speicher D.W. Marchesi V.T. J. Biol. Chem. 1986; 261: 3310-3315Abstract Full Text PDF PubMed Google Scholar, 4Horne W.C. Huang S.C. Becker P.S. Tang T.K. Benz Jr., E.J. Blood. 1993; 82: 2558-2563Crossref PubMed Google Scholar, 5Schischmanoff P.O. Winardi R. Discher D.E. Parra M.K. Bicknese S.E. Witkowska H.E. Conboy J.G. Mohandas N. J. Biol. Chem. 1995; 270: 21243-21250Crossref PubMed Scopus (57) Google Scholar) are associated with congenital hemolytic anemias (6Delaunay J. Crit. Rev. Oncol. Hematol. 1995; 19: 79-110Crossref PubMed Scopus (53) Google Scholar). Protein 4.1R also plays an important role in regulating the glycophorin C-4.1R-p55 ternary complex in the erythrocyte membrane (7Nunomura W. Takakuwa Y. Parra M. Conboy J.G. Mohandas N. J. Biol. Chem. 2000; 275: 6360-6367Abstract Full Text Full Text PDF PubMed Scopus (62) Google Scholar). Many immunological studies have shown that 4.1R protein is more complex in nonerythroid cells. Thus, 4.1R-immunoreactive polypeptides ranging in size from 30 to 210 kDa have been detected in different tissue and cell types (8Anderson R.A. Correas I. Mazzucco C. Castle J.D. Marchesi V.T. J. Cell. Biochem. 1988; 37: 269-284Crossref PubMed Scopus (47) Google Scholar, 9Granger B.L. Lazarides E. Cell. 1984; 37: 595-607Abstract Full Text PDF PubMed Scopus (85) Google Scholar) and 4.1R epitopes have been observed at many different sites, including stress fibers (10Cohen C.M. Foley S.F. Korsgren C. Nature. 1982; 299: 648-650Crossref PubMed Scopus (94) Google Scholar), centrosomes (11Krauss S.W. Chasis J.A. Rogers C. Mohandas N. Krockmalnic G. Penman S. Proc. Natl. Acad. Sci. U. S. A. 1997; 94: 7297-7302Crossref PubMed Scopus (52) Google Scholar), and the nucleus (12Correas I. Biochem. J. 1991; 279: 581-585Crossref PubMed Scopus (47) Google Scholar, 13De Cárcer G. Lallena M.J. Correas I. Biochem. J. 1995; 312: 871-877Crossref PubMed Scopus (69) Google Scholar, 14Krauss S.W. Larabell C.A. Lockett S. Gascard P. Penman S. Mohandas N. Chasis J.A. J. Cell Biol. 1997; 137: 275-289Crossref PubMed Scopus (96) Google Scholar, 15Lallena M.J. Correas I. J. Cell Sci. 1997; 110: 239-247PubMed Google Scholar). The nuclear localization of specific isoforms of 4.1R has recently been confirmed by transfection experiments of 4.1R cDNAs isolated from erythroid (16Gascard P. Lee G. Coulombel L. Auffray I. Lum M. Parra M. Conboy J.G. Mohandas N. Chasis J.A. Blood. 1998; 92: 4404-4414Crossref PubMed Google Scholar) and nonerythroid human cells (17Mattagajasingh S.N. Huang S.C. Hartenstein J.S. Snyder M. Marchesi V.T. Benz E.J. J. Cell Biol. 1999; 145: 29-43Crossref PubMed Scopus (112) Google Scholar, 18Luque C.M. Correas I. J. Cell Sci. 2000; 113: 2485-2495PubMed Google Scholar, 19Luque C.M. Lallena M.J. Alonso M.A. Correas I. J. Biol. Chem. 1998; 273: 11643-11649Abstract Full Text Full Text PDF PubMed Scopus (39) Google Scholar). The prototypical erythroid protein 4.1R80 is, therefore, only one of many isoforms that are generated by a single gene, mainly by extensive alternative splicing of the 4.1R pre-mRNA (20Tang T.K. Leto T.L. Correas I. Alonso M.A. Marchesi V.T. Benz Jr., E.J. Proc. Natl. Acad. Sci. U. S. A. 1988; 85: 3713-3717Crossref PubMed Scopus (80) Google Scholar, 21Tang T.K. Qin Z. Leto T. Marchesi V.T. Benz Jr., E.J. J. Cell Biol. 1990; 110: 617-624Crossref PubMed Scopus (89) Google Scholar, 22Conboy J.G. Chan J.Y. Chasis J.A. Kan Y.W. Mohandas N. J. Biol. Chem. 1991; 266: 8273-8280Abstract Full Text PDF PubMed Google Scholar, 23Conboy J.G. Chan J. Mohandas N. Kan Y.W. Proc. Natl. Acad. Sci. U. S. A. 1988; 85: 9062-9065Crossref PubMed Scopus (97) Google Scholar). 4.1R80 protein is produced when 17 nucleotides 5′-upstream from exon 2 are spliced out, and translation is initiated at the downstream start site present in exon 4 (ATG2). The synthesis of isoforms, termed 4.1R135, containing up to 209 amino acids to the NH2 terminus of erythroid 4.1R80, occurs when the 17-nucleotide sequence containing the upstream ATG (ATG1) translation initiation codon is included. These isoforms are predominantly expressed in nonerythroid cells (20Tang T.K. Leto T.L. Correas I. Alonso M.A. Marchesi V.T. Benz Jr., E.J. Proc. Natl. Acad. Sci. U. S. A. 1988; 85: 3713-3717Crossref PubMed Scopus (80) Google Scholar, 21Tang T.K. Qin Z. Leto T. Marchesi V.T. Benz Jr., E.J. J. Cell Biol. 1990; 110: 617-624Crossref PubMed Scopus (89) Google Scholar). A third type of isoforms, termed 4.1R60, can be produced in erythroid and nonerythroid cells when both the 17-nucleotide sequence (containing the ATG1) and exon 4 (containing the ATG2) are spliced out, and translation is initiated from a third translation initiation site (ATG3) present in exon 8 (16Gascard P. Lee G. Coulombel L. Auffray I. Lum M. Parra M. Conboy J.G. Mohandas N. Chasis J.A. Blood. 1998; 92: 4404-4414Crossref PubMed Google Scholar, 18Luque C.M. Correas I. J. Cell Sci. 2000; 113: 2485-2495PubMed Google Scholar). Although the major functions of 4.1R80 protein have been extensively characterized in mature erythrocytes, the potential roles of 4.1R isoforms in nucleated cells have only begun to be characterized. It has been reported that 4.1R protein interacts with various splicing factors (15Lallena M.J. Correas I. J. Cell Sci. 1997; 110: 239-247PubMed Google Scholar, 24Lallena M.J. Martı́nez C. Valcárcel J. Correas I. J. Cell Sci. 1998; 111: 1963-1971Crossref PubMed Google Scholar); with pICln (25Tang C.J. Tang T.K. Blood. 1998; 92: 1442-1447Crossref PubMed Google Scholar), an integral chloride channel component which was recently shown to associate also with spliceosomal proteins (26Pu W.T. Krapivinsky G.B. Krapivinsky L. Clapham D.E. Mol. Cell. Biol. 1999; 19: 4113-4120Crossref PubMed Scopus (85) Google Scholar); with a novel centrosomal protein, termed CPAP (27Hung L.Y. Tang C.J. Tang T.K. Mol. Cell. Biol. 2000; 20: 7813-7825Crossref PubMed Scopus (141) Google Scholar); and with the nuclear mitotic apparatus protein (17Mattagajasingh S.N. Huang S.C. Hartenstein J.S. Snyder M. Marchesi V.T. Benz E.J. J. Cell Biol. 1999; 145: 29-43Crossref PubMed Scopus (112) Google Scholar). All of these observations indicate that, in nucleated cells, isoforms of 4.1R protein may play roles in organizing the nuclear architecture and mitotic spindle poles. These roles were not suspected from the initial studies, given that they were performed in anucleate, nondividing, human red blood cells. Interactions of 4.1R with other proteins have also been reported (28Nunomura W. Takakuwa Y. Tokimitsu R. Krauss S.W. Kawashima M. Mohandas N. J. Biol. Chem. 1997; 272: 30322-30328Crossref PubMed Scopus (116) Google Scholar, 29Hou C.L. Tang C. Roffler S.R. Tang T.K. Blood. 2000; 96: 747-753Crossref PubMed Google Scholar, 30Mattagajasingh S.N. Huang S.C. Hartenstein J.S. Benz Jr., E.J. J. Biol. Chem. 2000; 275: 30573-30585Abstract Full Text Full Text PDF PubMed Scopus (127) Google Scholar, 31Kontrogianni-Konstantopoulos A. Frye C.S. Benz Jr., E.J. Huang S.C. J. Biol. Chem. 2001; 276: 20679-20687Abstract Full Text Full Text PDF PubMed Scopus (32) Google Scholar, 32Pasternack G.R. Racusen R.H. Proc. Natl. Acad. Sci. U. S. A. 1989; 86: 9712-9716Crossref PubMed Scopus (34) Google Scholar, 33Correas I. Avila J. Biochem. J. 1988; 255: 217-221PubMed Google Scholar, 34Kontrogianni-Konstantopoulos A. Huang S.C. Benz Jr., E.J. Mol. Biol. Cell. 2000; 11: 3805-3817Crossref PubMed Scopus (39) Google Scholar), thus suggesting that 4.1R protein may be involved in many different events in nucleated cells. In an attempt to characterize further the functional roles and partners of nonerythroid 4.1R isoforms, we have analyzed 4.1R distribution in human T cells and observed that both endogenous and exogenous 4.1R proteins colocalized with the microtubule skeleton. In vivoand in vitro biochemical analyses confirmed an association between 4.1R and microtubules. We have determined that a region conserved in all 4.1R isoforms, previously designated by us as the “core region” (18Luque C.M. Correas I. J. Cell Sci. 2000; 113: 2485-2495PubMed Google Scholar), was involved in tubulin binding and that 22 amino acids containing leucine residues, organized as heptad repeats, were essential for the interaction. Our results indicate that both ATG1- and ATG2-translated 4.1R isoforms are able to associate with microtubules in interphase human T cells, suggesting the involvement of 4.1R in the microtubule architecture. The finding that ectopic expression of 4.1R in COS-7 cells resulted in disorganization of the microtubule architecture supports the hypothesis that the functional activity of protein 4.1R depends on the cell type in which the protein is expressed. The cell lines used in this study were human T lymphoid Jurkat and CEM cells and fibroblastic COS-7 cells. Jurkat and CEM cells were grown in tissue culture flasks in RPMI 1640 medium (Life Technologies, Inc.). COS-7 cells were grown on culture dishes or on glass coverslips in Dulbecco's modified Eagle's medium (Life Technologies, Inc.). Both media were supplemented with 1% glutamine, 10% (v/v) fetal calf serum (Life Technologies, Inc.), penicillin (50 units/ml), and streptomycin (50 units/ml). Cultures were maintained at 37 °C under a 5% CO2, 95% air humidified atmosphere. Transfection experiments were performed by electroporation using the Electro Cell Manipulator 600 (8BTX, San Diego, CA). Cells were processed 48 h after transfection. The 4.1R cDNAs used for transfections (pSRα4.1R135Δ16; pCR3.1–4.1R135Δ16,19; pCR3.1–4.1R135Δ4,5,16; pSRα4.1R80Δ16 and pCR3.1–4.1R80Δ16,19), were isolated from MOLT-4 T cells as described previously (18Luque C.M. Correas I. J. Cell Sci. 2000; 113: 2485-2495PubMed Google Scholar, 19Luque C.M. Lallena M.J. Alonso M.A. Correas I. J. Biol. Chem. 1998; 273: 11643-11649Abstract Full Text Full Text PDF PubMed Scopus (39) Google Scholar, 35Luque C.M. Lallena M.J. Pérez-Ferreiro C.M. de Isidro Y. De Cárcer G. Alonso M.A. Correas I. Proc. Natl. Acad. Sci. U. S. A. 1999; 96: 14925-14930Crossref PubMed Scopus (18) Google Scholar). GST1-4.1R80Δ16, GST-Cter and GST-core were constructed by polymerase chain reaction using pSRα4.1R80Δ16 as template. For GST-4.1R60Δ16,18, the template pCR3.1–4.1R60Δ16,18 was used. Appropriate sense and antisense primers containing the BglII and XhoI restriction sites at the 5′ and 3′ ends, respectively, were used for the amplification reactions. The amplified cDNAs were inserted into the BamHI and XhoI sites of pGEX-6P1 vector (Amersham Pharmacia Biotech) in-frame with the GST coding sequence. The GST-coreΔleu construct with a 66-nucleotide deletion (1403–1469) in exon 10 (GenBank™ accession no. M61733) (22Conboy J.G. Chan J.Y. Chasis J.A. Kan Y.W. Mohandas N. J. Biol. Chem. 1991; 266: 8273-8280Abstract Full Text PDF PubMed Google Scholar) was obtained by polymerase chain reaction using the GST-core region construct as the template and sense and antisense oligonucleotide primers annealing to the flanking regions of the sequence to be deleted. All cDNA constructs were sequenced as previously described (19Luque C.M. Lallena M.J. Alonso M.A. Correas I. J. Biol. Chem. 1998; 273: 11643-11649Abstract Full Text Full Text PDF PubMed Scopus (39) Google Scholar). The GST fusion proteins were overexpressed in Escherichia coli BL21 cells and purified by glutathione affinity chromatography (Amersham Pharmacia Biotech) using standard protocols. Subsequently the proteins were dialyzed against CSF-XB buffer (10 mm K-Hepes, pH 7.7, 50 mm sucrose, 100 mm KCl, 2 mmMgCl2, 0.1 mm CaCl2, and 5 mm EGTA) (36Wittmann T. Boleti H. Antony C. Karsenti E. Vernos I. J. Cell Biol. 1998; 143: 673-685Crossref PubMed Scopus (164) Google Scholar), frozen in liquid nitrogen, and stored at −70 °C. Anti-c-Myc monoclonal antibody 9E10 (37Evan G.I. Lewis G.K. Ramsay G. Bishop J.M. Mol. Cell. Biol. 1985; 5: 3610-3616Crossref PubMed Scopus (2166) Google Scholar) was obtained from the American Type Culture Collection. Anti-4.1R (10b) antibody was an affinity-purified polyclonal antibody generated as described previously (2Correas I. Speicher D.W. Marchesi V.T. J. Biol. Chem. 1986; 261: 13362-13366Abstract Full Text PDF PubMed Google Scholar). Anti-4.1R (762) was a polyclonal antibody raised against a synthetic peptide whose sequence is encoded by exon 2 (35Luque C.M. Lallena M.J. Pérez-Ferreiro C.M. de Isidro Y. De Cárcer G. Alonso M.A. Correas I. Proc. Natl. Acad. Sci. U. S. A. 1999; 96: 14925-14930Crossref PubMed Scopus (18) Google Scholar). The anti-centrosome antibody was a human autoimmune serum that strongly recognizes centrosomes in mammalian cells (38Dominguez J.E. Buendia B. Lopez-Otin C. Antony C. Karsenti E. Avila J. J. Cell Sci. 1994; 107: 601-611PubMed Google Scholar). Anti-α-tubulin antibody DM1A was a monoclonal antibody obtained from Sigma. Anti-GST antibody was a polyclonal antibody from Sigma. Fluorescence- and horseradish peroxidase-labeled antibodies were obtained from Southern Biotechnology Associates, Inc. (Birmingham, AL). Human T cells were incubated in flat-bottomed, 24-well plates in a final volume of 500 μl of complete medium on glass coverslips coated with polylysine at 1 mg/ml. Cells were fixed, permeabilized, and blocked as described (19Luque C.M. Lallena M.J. Alonso M.A. Correas I. J. Biol. Chem. 1998; 273: 11643-11649Abstract Full Text Full Text PDF PubMed Scopus (39) Google Scholar). Cells were incubated with the appropriate antibodies and processed as reported (13De Cárcer G. Lallena M.J. Correas I. Biochem. J. 1995; 312: 871-877Crossref PubMed Scopus (69) Google Scholar). As 9E10 and DM1A are both mouse monoclonal antibodies, in Figs. 3 and 8 antibody 10b was used 25-fold diluted, relative to concentrations used for detection of endogenous protein, to detect only exogenous epitope-tagged 4.1R proteins (19Luque C.M. Lallena M.J. Alonso M.A. Correas I. J. Biol. Chem. 1998; 273: 11643-11649Abstract Full Text Full Text PDF PubMed Scopus (39) Google Scholar). In Fig.2 B cells were fixed and extracted at the same time with 5% formalin (37% formaldehyde solution, Sigma) and 0.5% Triton X-100 in phosphate-buffered saline for 3 min at room temperature. Images were obtained using a Bio-Rad Radiance 2000 confocal laser microscope or a Zeiss epifluorescence microscope.Figure 8Exogenous expression of 4.1R induces microtubule disorganization in COS-7 cells. COS-7 cells were transfected with a 4.1R cDNA encoding isoform 4.1R80Δ16 and processed for double immunofluorescence with antibodies 10b (10b) and DM1A (DM1A) 48 h after transfection. Cells were analyzed by epifluorescence microscopy. The white arrow indicates a representative transfected cell presenting a disorganized microtubule network, which is no longer well focused at the centrosome. Untransfected cells show typical microtubules radiating from the centrosome.View Large Image Figure ViewerDownload (PPT)Figure 2Colocalization of endogenous 4.1R with the tubulin skeleton of T cells. A, CEM cells fixed with formalin in the absence of Triton X-100 were double-labeled with the anti-tubulin antibody DM1A (b and d) and the anti-4.1R antibodies 10b (a) or 762 (c). Areas inlower right (a–d) of eachpanel show enlargements of indicated areas. B, CEM cells fixed with formalin containing Triton X-100 were double-labeled with a human anti-centrosome antibody (b andd) and the anti-4.1R antibodies 10b (a) or 762 (c). The samples were analyzed by confocal microscopy.View Large Image Figure ViewerDownload (PPT) Human T cells were washed twice with phosphate-buffered saline and lysed in Laemmli solubilizing buffer (39Laemmli U.K. Nature. 1970; 227: 680-685Crossref PubMed Scopus (207461) Google Scholar). For immunoblot analysis, protein fractions were separated by SDS-polyacrylamide gel electrophoresis and transferred to Immobilon polyvinylidene difluoride (Millipore) in Tris borate buffer, pH 8.2. Membranes were processed and developed as described (13De Cárcer G. Lallena M.J. Correas I. Biochem. J. 1995; 312: 871-877Crossref PubMed Scopus (69) Google Scholar). CEM cells were harvested by centrifugation and then resuspended in 1/10 of buffer A (0.1 m MES, 0.5 mm MgCl2, 2 mm EGTA) containing 10 μg/ml pepstatin, leupeptin, aprotinin, and 1 mm phenylmethylsulfonyl fluoride. Cells were lysed in 4 °C hypotonic buffer by 25 passages through a 22G1-gauge needle fitted onto a plastic syringe. 10× buffer A was added to the extract to obtain isotonic buffer A. The lysate was centrifuged at high speed in a minicentrifuge at 4 °C and processed as described (40Charrasse S. Schroeder M. Gauthier-Rouviere C. Ango F. Cassimeris L. Gard D.L. Larroque C. J. Cell Sci. 1998; 111: 1371-1383Crossref PubMed Google Scholar). Briefly, the pellet was discarded and the supernatant was centrifuged at 100,000 × g for 60 min at 4 °C in a Beckman TL-100 Tabletop Ultracentrifuge using a TLA-100.1 fixed-angle rotor. The pellet, corresponding to the membrane-containing fraction, was discarded. The supernatant was supplemented with 10 μm Taxol and 0.1 mm GTP and incubated for 35 min at 37 °C. Microtubules were centrifuged through a 15% sucrose cushion in buffer A containing 5 μm Taxol (30,000 × g, 35 min, 37 °C). The microtubule pellets and supernatants were boiled in Laemmli buffer and analyzed by Western blotting with 9E10, 10b, or 762 antibodies. PC-tubulin from bovine brain (41Weingarten M.D. Lockwood A.H. Hwo S.Y. Kirschner M.W. Proc. Natl. Acad. Sci. U. S. A. 1975; 72: 1858-1862Crossref PubMed Scopus (2230) Google Scholar) was polymerized as described (36Wittmann T. Boleti H. Antony C. Karsenti E. Vernos I. J. Cell Biol. 1998; 143: 673-685Crossref PubMed Scopus (164) Google Scholar), mixed with recombinant proteins in buffer BRB80 (80 mm K-Pipes, pH 6.8, 1 mmEGTA, 1 mm MgCl2) containing 10 μm Taxol, and incubated for 15 min at room temperature. The samples were centrifuged through a 15% sucrose cushion in BRB80, as described previously (36Wittmann T. Boleti H. Antony C. Karsenti E. Vernos I. J. Cell Biol. 1998; 143: 673-685Crossref PubMed Scopus (164) Google Scholar). Equivalent aliquots of supernatants and pellets were analyzed on Coomassie-stained gels and Western blots probed with anti-GST antibody. To understand how 4.1R is distributed in human T cells, we analyzed CEM cells, fixed in the absence (Fig. 2 A) or in the presence (Fig. 2 B) of Triton X-100, by confocal microscopy. Fig. 2 A (a) shows a representative image of cells stained with the anti-4.1R 10b antibody. Diffuse cytoplasmic staining, which was more concentrated at the pericentrosomal region, and staining of both discrete cytoplasmic filaments and the plasma membrane were observed. Nuclear staining (not shown in this confocal plane) was also detected. The cytoplasmic filaments decorated with antibody 10b corresponded to microtubules, as revealed by the anti-tubulin DM1A antibody (Fig. 2 A, b). Microtubules were also decorated with the anti-4.1R 762 antibody, which recognizes the extra amino-terminal region of 4.1R135 isoforms (35Luque C.M. Lallena M.J. Pérez-Ferreiro C.M. de Isidro Y. De Cárcer G. Alonso M.A. Correas I. Proc. Natl. Acad. Sci. U. S. A. 1999; 96: 14925-14930Crossref PubMed Scopus (18) Google Scholar) (Fig.2 A, c and d). These results indicated that the microtubule network of human T cells contains 4.1R epitopes. The 4.1R staining pattern of CEM cells fixed in the presence of Triton X-100 is shown in Fig. 2 B. The Triton X-100 treatment resulted in the extraction of most of the 4.1R immunoreactivity; however, nuclear speckles and a bright spot in the pericentrosomal region were clearly observed in cells stained with the 10b antibody (Fig. 2 B, a). The bright spot was stained with antibodies that recognize the centrosome (Fig. 2 B,b). Antibody 762 did not stain the centrosome (Fig.2 B, c). These results show that centrosomal 4.1R immunoreactivity is better detected when cells are fixed in the presence of Triton X-100. To identify specific 4.1R isoforms that colocalize with microtubules, we transfected T cells with different 4.1R cDNAs previously isolated by us from human T cells (18Luque C.M. Correas I. J. Cell Sci. 2000; 113: 2485-2495PubMed Google Scholar, 35Luque C.M. Lallena M.J. Pérez-Ferreiro C.M. de Isidro Y. De Cárcer G. Alonso M.A. Correas I. Proc. Natl. Acad. Sci. U. S. A. 1999; 96: 14925-14930Crossref PubMed Scopus (18) Google Scholar) and compared the intracellular distribution of the expressed 4.1R proteins and that of tubulin by confocal microscopy. Two of the 4.1R cDNAs (4.1R80Δ16 and 4.1R80Δ16,19) encode 4.1R isoforms that are translated from the ATG-2 translation initiation codon present in exon 4 (Fig.1 B), whereas the other three 4.1R cDNAs (4.1R135Δ16, 4.1R135Δ16,19, and 4.1R135Δ4,5,16) encode 4.1R isoforms that are translated from the 5′ upstream-ATG-1 translation initiation site (Fig.1 B). The only difference between the isoforms 4.1R135Δ16 and 4.1R135Δ16,19 and their respective counterparts 4.1R80Δ16 and 4.1R80Δ16,19 is that the first two have the 209-amino acid NH2-terminal extension. Fig. 3 shows confocal microscopy images of Jurkat T cells transiently expressing the 4.1R isoforms and double-stained with 10b (10b) and DM1A (DM1A) antibodies. Antibody 10b was highly diluted to react only with the exogenously expressed 4.1R isoforms but not with the endogenous 4.1R proteins. All 4.1R isoforms localized to cytoplasmic filaments (Fig. 3, A(a, c, and e) and B(a and c)) that were also stained by the tubulin-recognizing monoclonal antibody DM1A (Fig. 3 A(b, d, and f) and B(b and d)). The distribution of the exogenously expressed 4.1R isoforms on the microtubule network resembled that of the endogenous 4.1R proteins (compare Figs.2 and3). To determine whether 4.1R and microtubules interacted in vivo, we performed biochemical assays using T cells stably transfected with either 4.1R135Δ16 or 4.1R80Δ16 cDNAs. Taxol-polymerized microtubules were isolated from these cells by centrifugation through a sucrose cushion, and the presence or absence of the expressed 4.1R proteins in the pellet fractions was analyzed by Western blot. Control cells, transfected with an empty plasmid, were processed in parallel. Expression of 4.1R135Δ16 and 4.1R80Δ16 in CEM cells was confirmed by Western blot analysis revealed with antibody 9E10 (Fig. 4 A), as the exogenous 4.1R isoforms were tagged at their COOH-terminal region with the 9E10 c-Myc epitope to distinguish them from the endogenous 4.1R proteins (19Luque C.M. Lallena M.J. Alonso M.A. Correas I. J. Biol. Chem. 1998; 273: 11643-11649Abstract Full Text Full Text PDF PubMed Scopus (39) Google Scholar). Western blots of equivalent aliquots of microtubule pellets and supernatants isolated from control and transfected cells were revealed with 9E10 antibody (Fig. 4 B). Isoforms 4.1R135Δ16 and 4.1R80Δ16 were present in the microtubule pellets (Fig. 4 B, lanes 3 and5), suggesting their in vivo association with the tubulin skeleton. Duplicates of the microtubule pellet shown in Fig. 4 B(lane 1) were revealed with anti-4.1R antibodies to analyze endogenous 4.1R proteins cosedimenting in the microtubule pellet (Fig.4 C). Antibody 10b (Fig. 4 C, 10b) detected major bands of ∼80 and 145 kDa. Antibody 762 (Fig.4 C, 762) reacted with a 145-kDa band. These results and those obtained for the exogenously expressed 4.1R isoforms indicate that both ATG1- and ATG2-translated 4.1R isoforms cosediment with microtubules. To investigate whether 4.1R80Δ16 could bind directly to tubulin, the major component of microtubules, we prepared tubulin depleted in microtubule-associated proteins (PC-tubulin) and performed in vitro binding assays. Taxol-polymerized tubulin was incubated with a GST fusion protein containing 4.1R80Δ16 (GST-4.1R80Δ16) and processed as indicated under “Experimental Procedures.” Protein GST-4.1R80Δ16 was detected in the pellet fraction with polymerized tubulin (Fig. 5 A,lane 1) but not in the supernatant fraction (Fig.5 A, lane 5). Similar results were confirmed by the blot analysis (Fig. 5 B, lanes 1 and5). To identify the 4.1R80Δ16 region interacting with tubulin, we fused various 4.1R fragments to GST (Fig.1 C) and assayed in vitro their ability to associate with PC-tubulin. Fig. 5 C shows a Coomassie-stained gel of the different GST fusion proteins used for the study. Protein GST-Cter (Fig. 5 C, lane 3) contained the carboxyl-terminal region of protein 4.1R80Δ16 (see Fig.1 C), and the GST-4.1R60Δ16,18 protein (Fig.5 C, lane 2) contained a short 4.1R isoform translated from the ATG3 triplet (18Luque C.M. Correas I. J. Cell Sci. 2000; 113: 2485-2495PubMed Google Scholar) whose sequence is comprised in 4.1R80Δ16 (see Fig. 1 C). Fig. 5(panels A (Coomassie-stained gels) andB (Western blots revealed with anti-GST)) shows that GST-4.1R60Δ16,18 protein was detected in the tubulin pellets (Fig. 5, A and B, lanes 2) but not in the supernatants (Fig. 5, A and B,lanes 6). By contrast, protein GST-Cter, assayed at two different concentrations, was observed in the supernatant fractions (Fig. 5, A and B, lanes 7 and8). The absence of GST-Cter protein from the tubulin pellets was confirmed by Western blot analysis (Fig. 5 B, lanes 3 and 4). The fact that GST-Cter did not bind to tubulin, whereas the GST-4.1R60Δ16,18 and GST-4.1R80Δ16 proteins did, suggested that the common central region of the latter proteins, previously designated by us as “the core region” (18Luque C.M. Correas I. J. Cell Sci. 2000; 113: 2485-2495PubMed Google Scholar), was responsible for their association with tubulin. A GST fusion protein containing the core region (GST-core) was assayed for its tubulin binding ability. Fig.6 A shows the Coomassie-stained gel of the pellet and the supernatant fractions and Fig. 6 Bthe Western blot revealed with anti-GST antibody. Protein GST-core was present in the pellet with tubulin (Fig. 6, A andB, lanes 3) and absent in the supernatant (Fig.6, A and B, lanes 6). The fusion proteins GST-4.1R80Δ16 (Fig. 6, A andB, lanes 1 and 4) and GST-Cter (Fig.6, A and B, lanes 2 and 5) were used as positive and negative controls, respectively. These results confirmed that the core region, present in all 4.1R isoforms, is responsible for tubulin binding. Primary sequence analysis of the core region revealed that the NH2-terminal region of exon 10-encoded sequences comprised heptad repeats of leucine residues resembling a putative leucine zipper motif (Fig.7 A). To investigate whether this characteristic sequence was involved in tubulin association, we created a deletion mutant, termed core-Δleu, that lacked 22 amino acids from the leucine zipper-resembling motif (amino acidsboxed in Fig. 7 A), fused it to GST (GST-core-Δleu) (see Fig. 1 C), and determined its tubulin-binding capacity. As expected, GST-core was detected in the pellet fraction (Fig. 7, B and C, lanes 1), whereas GST-core-Δleu remained in the supernatant (Fig. 7,B and C, lanes 4), indicating that the 22-amino acid stretch containing the heptad repeats of leucine is essential for tubulin binding. COS-7 cells were transfected with 4.1R80Δ16 cDNA to compare the distribution pattern of the expressed 4.1R isoform with that of tubulin by double immunofluorescence microscopy (Fig.8). Isoform 4.1R80Δ16 did not result to colocalize with microtubules as it did in T cells but instead led to disorganization of the microtubule network (compare Fig.8 with Fig. 3 B, a and b). Similar disorganization of the microtubule architecture was also observed with other 4.1R cDNAs assayed (data not shown). A representative image of the altered microtubule network showing microtubules that no longer radiate from a single perinuclear focus is represented in Fig. 8. The great diversity of 4.1R isoforms present in nonerythroid cells makes it necessary to assay individual isoforms to specifically assign them cellular functions. In recent years, much effort has been concentrated on isolating 4.1R cDNAs from different cell sources for use in the assignment of roles and partners for specific 4.1R isoforms. In this study we show that naturally occurring exogenously expressed ATG1- and ATG2-translated 4.1R isoforms, and also endogenous 4.1R proteins, colocalize with the tubulin cytoskeleton of interphase human T cells. Cosedimentation of 4.1R proteins with microtubules of human T cells and direct in vitro binding to tubulin revealed that a common region present in all 4.1R isoforms, the core region, is responsible for the association of 4.1R with tubulin. This is the first demonstration of an association between 4.1R and interphase microtubules. It is of particular note that the 22-amino acid sequence required for 4.1R-tubulin interaction contains heptad repeats of leucine resembling leucine zippers (see Fig. 7 A) but, unlike the latter, it does not adopt an α-helix conformation (42O'Shea E.K. Rutkowski R. Kim P.S. Science. 1989; 243: 538-542Crossref PubMed Scopus (698) Google Scholar). The crystal structure of the NH2-terminal 30-kDa domain of 4.1R has been determined and has the form of a three-lobed cloverleaf (43Han B.G. Nunomura W. Takakuwa Y. Mohandas N. Jap B.K. Nat. Struct. Biol. 2000; 7: 871-875Crossref PubMed Scopus (108) Google Scholar). The COOH-terminal lobe contains two β-sheets and ends in an α-helix; one of the β-sheets contains the binding site for p55, a palmitoylated peripheral membrane protein belonging to a membrane-associated guanylate kinase homologue family of signaling and cytoskeletal proteins (7Nunomura W. Takakuwa Y. Parra M. Conboy J.G. Mohandas N. J. Biol. Chem. 2000; 275: 6360-6367Abstract Full Text Full Text PDF PubMed Scopus (62) Google Scholar, 44Marfatia S.M. Leu R.A. Branton D. Chishti A.H. J. Biol. Chem. 1995; 270: 715-719Abstract Full Text Full Text PDF PubMed Scopus (178) Google Scholar). The 22 amino acids identified in this study as being required for 4.1R-tubulin associations would be located on the other β-sheet of the COOH-terminal lobe. Removal of the 22 amino acids would result in an almost complete absence of this β-sheet structure. Ezrin-radixin-moesin (ERM) proteins belong to the “4.1R protein family” and are thought to link actin filaments to the plasma membrane at cortical foci (45Bretscher A. Curr. Opin. Cell Biol. 1999; 11: 109-116Crossref PubMed Scopus (331) Google Scholar). Ezrin has also been isolated from microtubule-associated protein preparations from Madin-Darby canine kidney and A72 cells and the ERM protein merlin contains a “cryptic” microtubule binding site exposed specifically in the activated or “open” ERM conformation (reviewed in Ref. 45Bretscher A. Curr. Opin. Cell Biol. 1999; 11: 109-116Crossref PubMed Scopus (331) Google Scholar). Thus, it has been suggested that some ERM proteins may play an additional role in linking microtubules to the cell cortex, thus having the capacity to associate with the microtubule and actin cytoskeletons. One of the major functions of erythroid 4.1R is the stabilization of the spectrin-actin complex through the 10-kDa domain (2Correas I. Speicher D.W. Marchesi V.T. J. Biol. Chem. 1986; 261: 13362-13366Abstract Full Text PDF PubMed Google Scholar, 3Correas I. Leto T.L. Speicher D.W. Marchesi V.T. J. Biol. Chem. 1986; 261: 3310-3315Abstract Full Text PDF PubMed Google Scholar, 4Horne W.C. Huang S.C. Becker P.S. Tang T.K. Benz Jr., E.J. Blood. 1993; 82: 2558-2563Crossref PubMed Google Scholar, 5Schischmanoff P.O. Winardi R. Discher D.E. Parra M.K. Bicknese S.E. Witkowska H.E. Conboy J.G. Mohandas N. J. Biol. Chem. 1995; 270: 21243-21250Crossref PubMed Scopus (57) Google Scholar) encoded by exons 16 and 17. We may speculate that, in nonerythroid cells, 4.1R isoforms expressing the alternative exon 16 would have the capacity to bind to the actin and the microtubule cytoskeletons, whereas those 4.1R isoforms lacking exon 16 expression would only retain the ability to bind to the microtubule cytoskeleton. It has been indicated that nonerythroid 4.1R isoforms may have different functional activities, depending on the cell type in which they are expressed (31Kontrogianni-Konstantopoulos A. Frye C.S. Benz Jr., E.J. Huang S.C. J. Biol. Chem. 2001; 276: 20679-20687Abstract Full Text Full Text PDF PubMed Scopus (32) Google Scholar). Thus, 4.1R interacts through its spectrin-actin binding domain with a protein complex in skeletal muscle (34Kontrogianni-Konstantopoulos A. Huang S.C. Benz Jr., E.J. Mol. Biol. Cell. 2000; 11: 3805-3817Crossref PubMed Scopus (39) Google Scholar) that differs from that described in PC12 cells (31Kontrogianni-Konstantopoulos A. Frye C.S. Benz Jr., E.J. Huang S.C. J. Biol. Chem. 2001; 276: 20679-20687Abstract Full Text Full Text PDF PubMed Scopus (32) Google Scholar), even though the latter type of cell also contains the proteins to which 4.1R binds in skeletal muscle. Concomitantly, we show in this study that transfection of T cells with T cell-4.1R cDNAs resulted in overexpression of 4.1R isoforms colocalizing with the microtubule network, whereas ectopic expression of 4.1R isoforms in COS-7 cells did not result in 4.1R binding to the microtubules but instead in the disruption of the microtubule architecture. It is evident, therefore, that different cell types respond in distinct manners to the expression of 4.1R, supporting the hypothesis that some functional activities of protein 4.1R are cell type-regulated. A major difference between T and COS-7 cells is that T cells experiment internal protein rearrangements during polarization in response to many stimuli (see below). Whether some cell type-specific partners of 4.1R are required for specialized 4.1R roles remains to be established. Polarization is a key feature in the biology of T cells, as T lymphocytes acquire a polarized phenotype after activation, upon interaction with antigen-presenting cells, and during transendothelial migration. To extravasate, circulating lymphocytes must adopt a polarized flexible form suitable for tissue invasion. The anterior region of the polarized lymphocyte bears multiple, highly labile lamellipodia, whereas the posterior part is drawn out into a single slender appendage called the uropod (reviewed in Ref. 46Sanchez-Madrid F. del Pozo M.A. EMBO J. 1999; 18: 501-511Crossref PubMed Scopus (523) Google Scholar). Polarization involves a reorganization of the cytoskeleton, including polymerization and redistribution of actin and a drastic reconfiguration of the tubulin cytoskeleton, which, in conjunction with the microtubule organizing center, retract into the uropod lumen, collapsing into a thin, compact sheaf. Microtubule retraction has been suggested as being a strategy for accelerating extravasation without disassembling the microtubule-based transport system (47Ratner S. Sherrod W.S. Lichlyter D. J. Immunol. 1997; 159: 1063-1067PubMed Google Scholar). The distribution of spectrin has been analyzed during lymphocyte activation, revealing a rapid reorganization of the protein, whereby the initial aggregated cytoplasmic structures are translocated to specific areas of the plasma membrane (48Lee J.K. Black J.D. Repasky E.A. Kubo R.T. Bankert R.B. Cell. 1988; 55: 807-816Abstract Full Text PDF PubMed Scopus (47) Google Scholar). Different distributions are observed for proteins of the ERM family in T lymphocytes induced to polarize by chemokines. Thus, radixin colocalizes with myosin II in the neck of the uropod, whereas moesin is preferentially located at the uropod tip, interacting with the cytoplasmic tail of ICAM-3, CD43, and CD44 (49Serrador J.M. Alonso-Lebrero J.L. del Pozo M.A. Furthmayr H. Schwartz-Albiez R. Calvo J. Lozano F. Sanchez-Madrid F. J. Cell Biol. 1997; 138: 1409-1423Crossref PubMed Scopus (202) Google Scholar). Interactions between 4.1R and CD44 have also been reported (28Nunomura W. Takakuwa Y. Tokimitsu R. Krauss S.W. Kawashima M. Mohandas N. J. Biol. Chem. 1997; 272: 30322-30328Crossref PubMed Scopus (116) Google Scholar). Further studies will be conducted to determine the behavior of 4.1R during T cell polarization, which may provide new clues for understanding the complex cytoskeletal reorganization experienced by this cell type. We are very grateful to Dr. Isabelle Vernos for invaluable help. We also thank Drs. Jesús Avila and Jorge Domı́nguez for providing us with PC-tubulin and anti-centrosome antibody, and Drs. Miguel A. Alonso, Jaime Millán, and Antonio Rodrı́guez for generous comments. We acknowledge Dr. Carlos Sánchez for assistance with confocal microscopy techniques.
Referência(s)