Structural Aspects of the Association of FcεRI with Detergent-resistant Membranes
1999; Elsevier BV; Volume: 274; Issue: 3 Linguagem: Inglês
10.1074/jbc.274.3.1753
ISSN1083-351X
AutoresKenneth A. Field, David Holowka, Barbara Baird,
Tópico(s)Cell Adhesion Molecules Research
ResumoWe recently showed that aggregation of the high affinity IgE receptor on mast cells, FcεRI, causes this immunoreceptor to associate rapidly with specialized regions of the plasma membrane, where it is phosphorylated by the tyrosine kinase Lyn. In this study, we further characterize the detergent sensitivity of this association on rat basophilic leukemia-2H3 mast cells, and we compare the capacity of structural variants of FcεRI and other receptors to undergo this association. We show that this interaction is not mediated by the β subunit of the receptor or the cytoplasmic tail of the γ subunit, both of which are involved in signaling. Using chimeric receptor constructs, we found that the extracellular segment of the FcεRI α subunit was not sufficient to mediate this association, implicating FcεRI α and/or γ transmembrane segments. To determine the specificity of this interaction, we compared the association of several other receptors. Interleukin-1 type I receptors on Chinese hamster ovary cells and α4 integrins on rat basophilic leukemia cells showed little or no association with isolated membrane domains, both before and after aggregation on the cells. In contrast, interleukin-2 receptor α (Tac) on Chinese hamster ovary cells exhibited aggregation-dependent membrane domain association similar to FcεRI. These results provide insights into the structural basis and selectivity of lipid-mediated interactions between certain transmembrane receptors and detergent-resistant membranes. We recently showed that aggregation of the high affinity IgE receptor on mast cells, FcεRI, causes this immunoreceptor to associate rapidly with specialized regions of the plasma membrane, where it is phosphorylated by the tyrosine kinase Lyn. In this study, we further characterize the detergent sensitivity of this association on rat basophilic leukemia-2H3 mast cells, and we compare the capacity of structural variants of FcεRI and other receptors to undergo this association. We show that this interaction is not mediated by the β subunit of the receptor or the cytoplasmic tail of the γ subunit, both of which are involved in signaling. Using chimeric receptor constructs, we found that the extracellular segment of the FcεRI α subunit was not sufficient to mediate this association, implicating FcεRI α and/or γ transmembrane segments. To determine the specificity of this interaction, we compared the association of several other receptors. Interleukin-1 type I receptors on Chinese hamster ovary cells and α4 integrins on rat basophilic leukemia cells showed little or no association with isolated membrane domains, both before and after aggregation on the cells. In contrast, interleukin-2 receptor α (Tac) on Chinese hamster ovary cells exhibited aggregation-dependent membrane domain association similar to FcεRI. These results provide insights into the structural basis and selectivity of lipid-mediated interactions between certain transmembrane receptors and detergent-resistant membranes. Multichain immune recognition receptors present on hematopoietic cells interact with Src family protein tyrosine kinases (PTKs) 1The abbreviations used are: PTK, protein tyrosine kinase; ITAM, immunoreceptor tyrosine-based activation motif; RBL, rat basophilic leukemia; DRM, detergent-resistant membrane; GPI, glycosylphosphatidylinositol; CHO, Chinese hamster ovary; IL, interleukin; TX-100, Triton X-100. 1The abbreviations used are: PTK, protein tyrosine kinase; ITAM, immunoreceptor tyrosine-based activation motif; RBL, rat basophilic leukemia; DRM, detergent-resistant membrane; GPI, glycosylphosphatidylinositol; CHO, Chinese hamster ovary; IL, interleukin; TX-100, Triton X-100. as an early signaling step (1Cambier J.C. J. Immunol. 1995; 155: 3281-3285PubMed Google Scholar). Aggregation of the high affinity receptor for IgE, FcεRI, results in phosphorylation of the β and γ2 subunits of this receptor by the Src family PTK, Lyn (2Scharenberg A.M. Lin S. Cuenod B. Yamamura H. Kinet J.P. EMBO J. 1995; 14: 3385-3394Crossref PubMed Scopus (143) Google Scholar, 3Zhang J. Berenstein E.H. Evans R.L. Siraganian R.P. J. Exp. Med. 1996; 184: 71-79Crossref PubMed Scopus (238) Google Scholar). Lyn phosphorylation of the immunoreceptor tyrosine-based activation motif (ITAM) within FcεRI γ subunits allows the ZAP-70-related PTK, Syk, to associate with the receptor via its two Src homology 2 domains (1Cambier J.C. J. Immunol. 1995; 155: 3281-3285PubMed Google Scholar, 2Scharenberg A.M. Lin S. Cuenod B. Yamamura H. Kinet J.P. EMBO J. 1995; 14: 3385-3394Crossref PubMed Scopus (143) Google Scholar). This recruitment and consequent activation of Syk leads to further downstream signaling, including phosphorylation and activation of phospholipase Cγ, mobilization of intracellular calcium, and activation of protein kinase C (4Beaven M.A. Baumgartner R.A Curr. Opin. Immunol. 1996; 8: 766-772Crossref PubMed Scopus (76) Google Scholar). In mast cells and basophils, which both express FcεRI, these cascades result in the exocytosis of preformed granules containing histamine and other vasoactive compounds, as well as other cellular responses. The phosphorylation of FcεRI by Lyn is a critical event in receptor activation, but the mechanism by which receptor aggregation stimulates this event is not well understood. We have proposed a novel model for this process in which specialized membrane domains enriched in Lyn mediate this phosphorylation that occurs after the aggregation-dependent association of FcεRI with these domains (5Field K.A. Holowka D. Baird B. Razin E. Rivera J. Signal Transduction in Mast Cells and Basophils. Springer-Verlag, New York1999: 102-114Crossref Google Scholar). This model is consistent with our findings that the rapid association of FcεRI with these domains does not depend on receptor phosphorylation (6Field K.A. Holowka D. Baird B. J. Biol. Chem. 1997; 272: 4276-4280Abstract Full Text Full Text PDF PubMed Scopus (325) Google Scholar). It is strongly supported by preferential tyrosine phosphorylation of those receptors associated with membrane domains. Moreover, in vitro tyrosine kinase assays reproduce this activation step within these membrane domains that can be isolated because of their resistance to detergent solubilization (6Field K.A. Holowka D. Baird B. J. Biol. Chem. 1997; 272: 4276-4280Abstract Full Text Full Text PDF PubMed Scopus (325) Google Scholar). Other models that require a direct interaction between Lyn and the β subunit of monomeric FcεRI (7Pribluda V.S. Pribluda C. Metzger H. Proc. Natl. Acad. Sci. U. S. A. 1994; 91: 11246-11250Crossref PubMed Scopus (173) Google Scholar, 8Jouvin M.H. Adamczewski M. Numerof R. Letourneur O. Valle A. Kinet J.P. J. Biol. Chem. 1994; 269: 5918-5925Abstract Full Text PDF PubMed Google Scholar) do not explain some observations, including the capacity of mutant and chimeric receptors lacking the β subunit to activate cells (9Alber G. Miller L. Jelsema C.L. Varin-Blank N. Metzger H. J. Biol. Chem. 1991; 266: 22613-22620Abstract Full Text PDF PubMed Google Scholar, 10Lin S. Cicala C. Scharenberg A.M. Kinet J.P. Cell. 1996; 85: 985-995Abstract Full Text Full Text PDF PubMed Scopus (307) Google Scholar, 11Juergens M. Wollenberg A. Hanau D. Henri-De-La-Salle H. Bieber T. J. Immunol. 1995; 155: 5184-5189PubMed Google Scholar, 12Eiseman E. Bolen J.B. J. Biol. Chem. 1992; 267: 21027-21032Abstract Full Text PDF PubMed Google Scholar, 13Wilson B.S. Kapp N. Lee R.J. Pfeiffer J.R. Martinez A.M. Platt Y. Letourneur F. Oliver J.M. J. Biol. Chem. 1995; 270: 4013-4022Abstract Full Text Full Text PDF PubMed Scopus (63) Google Scholar, 14Repetto B. Bandara G. Kado-Fong H. Larigan J.D. Wiggan G.A. Pocius D. Basu M. Gilfillan A.M. Kochan J.P J. Immunol. 1996; 156: 4876-4883PubMed Google Scholar) or the difficulty in identifying the molecular basis of this interaction (15Vonakis B.M. Chen H. Haleem-Smith H. Metzger H. J. Biol. Chem. 1997; 272: 24072-24080Crossref PubMed Scopus (106) Google Scholar). Lyn association with the detergent-resistant membranes (DRMs) isolated from the RBL-2H3 mucosal mast cell line (16Field K.A. Holowka D. Baird B. Proc. Natl. Acad. Sci. U. S. A. 1995; 92: 9201-9205Crossref PubMed Scopus (271) Google Scholar) occurs as in other cells for Src family PTKs that contain a consensus site for dual fatty acid modifications (17Resh M.D. Cell. 1994; 76: 411-413Abstract Full Text PDF PubMed Scopus (592) Google Scholar, 18Shenoy-Scaria A.M. Dietzen D.J. Kwong J. Link D.C. Lublin D.M. J. Cell Biol. 1994; 126: 353-363Crossref PubMed Scopus (343) Google Scholar). These low density membranes are isolated by sucrose gradient ultracentrifugation of Triton X-100 (TX-100)-lysed cells. In other cell types, these preparations have been shown to be enriched in cholesterol, sphingomyelin, and gangliosides (19Brown D.A. Rose J.K. Cell. 1992; 68: 533-544Abstract Full Text PDF PubMed Scopus (2609) Google Scholar), as well as a subset of membrane-associated proteins, including certain Src family PTKs, heterotrimeric GTP-binding proteins, and glycosylphosphatidylinositol (GPI)-linked proteins (18Shenoy-Scaria A.M. Dietzen D.J. Kwong J. Link D.C. Lublin D.M. J. Cell Biol. 1994; 126: 353-363Crossref PubMed Scopus (343) Google Scholar, 19Brown D.A. Rose J.K. Cell. 1992; 68: 533-544Abstract Full Text PDF PubMed Scopus (2609) Google Scholar, 20Arreaza G. Melkonian K.A. LaFevre-Bernt M. Brown D.A. J. Biol. Chem. 1994; 269: 19123-19127Abstract Full Text PDF PubMed Google Scholar, 21Sargiacomo M. Sudol M. Tang Z. Lisanti M.P. J. Cell Biol. 1993; 122: 789-807Crossref PubMed Scopus (862) Google Scholar). Preparations similar to DRMs are also referred to as detergent-insoluble glycolipid-enriched domains (DIGs), glycolipid-enriched membranes (GEMs), or sphingolipid-cholesterol rafts (22Harder T. Simons K. Curr. Opin. Cell Biol. 1997; 9: 534-542Crossref PubMed Scopus (717) Google Scholar, 23Brown D.A. London I. Annu. Rev. Dev. Cell Biol. 1999; 14: 111-136Crossref Scopus (2549) Google Scholar). Caveolae, which are flask-shaped invaginations on the plasma membrane that contain the marker protein caveolin, can also be isolated using similar methods and appear to contain many of the same components (24Anderson R.G.W. Proc. Natl. Acad. Sci. U. S. A. 1993; 90: 10909-10913Crossref PubMed Scopus (569) Google Scholar, 25Lisanti M.P. Scherer P.E. Vidugiriene J. Tang Z. Hermanowski-Vosatka A. Tu Y.H. Cook R.F. Sargiacomo M. J. Cell Biol. 1994; 126: 111-126Crossref PubMed Scopus (813) Google Scholar). RBL cells appear to be like other hematopoietic cells, including T- and B-cell lines (26Gorodinsky A. Harris D.A. J. Cell Biol. 1995; 129: 619-627Crossref PubMed Scopus (296) Google Scholar, 27Fra A.M. Williamson E. Simons K. Parton R.G. J. Biol. Chem. 1994; 269: 30745-30748Abstract Full Text PDF PubMed Google Scholar, 28Parolini I. Sargiacomo M. Lisanti M.P. Peschle C. Blood. 1996; 87: 3783-3794Crossref PubMed Google Scholar, 29Fra A.M. Williamson E. Simons K. Parton R.G. Proc. Natl. Acad. Sci. U. S. A. 1995; 92: 8655-8659Crossref PubMed Scopus (525) Google Scholar), that do not contain caveolae but do exhibit DRMs that are enriched in signaling molecules (16Field K.A. Holowka D. Baird B. Proc. Natl. Acad. Sci. U. S. A. 1995; 92: 9201-9205Crossref PubMed Scopus (271) Google Scholar). FcεRI associates with isolated DRMs when the receptor is aggregated at the cell surface (6Field K.A. Holowka D. Baird B. J. Biol. Chem. 1997; 272: 4276-4280Abstract Full Text Full Text PDF PubMed Scopus (325) Google Scholar). Preservation of this association after cell lysis depends on the stability of the receptor aggregate and on using a low concentration of TX-100 in the lysate (6Field K.A. Holowka D. Baird B. J. Biol. Chem. 1997; 272: 4276-4280Abstract Full Text Full Text PDF PubMed Scopus (325) Google Scholar, 16Field K.A. Holowka D. Baird B. Proc. Natl. Acad. Sci. U. S. A. 1995; 92: 9201-9205Crossref PubMed Scopus (271) Google Scholar). These interactions can be detected at the intact cell surface, as observed with fluorescence microscopy. For example, aggregation of FcεRI into patches causes co-redistribution of DiI-C16, a fluorescent lipid probe with saturated acyl chains, and this lipid analog has reduced lateral mobility in these patches (30Thomas J.L. Holowka D. Baird B. Webb W.W. J. Cell Biol. 1994; 125: 795-802Crossref PubMed Scopus (135) Google Scholar). Co-redistribution with aggregated FcεRI on intact cells is also observed for three other membrane components isolated with DRMs (16Field K.A. Holowka D. Baird B. Proc. Natl. Acad. Sci. U. S. A. 1995; 92: 9201-9205Crossref PubMed Scopus (271) Google Scholar), a GD1bganglioside derivative (31Pierini L. Holowka D. Baird B. J. Cell Biol. 1996; 134: 1427-1439Crossref PubMed Scopus (65) Google Scholar), the GPI-linked protein Thy-1, 2D. Holowka, unpublished observations. 2D. Holowka, unpublished observations.and Lyn.2 In the present study, we investigated the structural basis for the interaction of FcεRI, a multisubunit transmembrane receptor, with isolated DRMs. These membranes were visualized by whole-mount electron microscopy to compare them with similar preparations from other cells. We also compared the association of wild-type and mutant FcεRI with DRMs isolated from hematopoietic and nonhematopoietic cell lines. Finally, we investigated the specificity of this interaction by measuring the aggregation-dependent association of other transmembrane receptors with DRMs. These results provide evidence that the structural features of FcεRI that mediate the detergent-sensitive interaction with membrane domains occur selectively but not uniquely with this receptor. RBL-2H3 cells, mouse monoclonal anti-dinitrophenyl IgE, and biotinylated 125I-IgE were previously described (6Field K.A. Holowka D. Baird B. J. Biol. Chem. 1997; 272: 4276-4280Abstract Full Text Full Text PDF PubMed Scopus (325) Google Scholar, 16Field K.A. Holowka D. Baird B. Proc. Natl. Acad. Sci. U. S. A. 1995; 92: 9201-9205Crossref PubMed Scopus (271) Google Scholar). P815 mouse mastocytoma cells and Chinese hamster ovary (CHO) cells stably transfected with wild-type or mutant FcεRI were generously provided by Dr. H. Metzger (National Institutes of Health) and were maintained as described (32Mao S.Y. Alber G. Rivera J. Kochan J. Metzger H. Proc. Natl. Acad. Sci. U. S. A. 1992; 89: 222-226Crossref PubMed Scopus (27) Google Scholar). Prior to harvesting, cells were sensitized with biotinylated 125I-IgE for 4–24 h. All stable FcεRI transfectants expressed at least 50,000 IgE receptors per cell at the time of the experiments except the αγ2 P815 cell line, which expressed approximately 18,000 receptors per cell. CHO cells stably transfected with type I IL-1 receptors were from Dr. S. Dower (University of Sheffield, United Kingdom) and were maintained as described (33Guo C. Dower S.K. Holowka D. Baird B. J. Biol. Chem. 1995; 270: 27562-27568Abstract Full Text Full Text PDF PubMed Scopus (47) Google Scholar). The IL-2 receptor α (Tac) and Chimera-1 receptor were transiently expressed on CHO cells for 48 h prior to performing experiments. The IL-2 receptor α DNA was provided in a pCMV mammalian expression vector by Dr. B. Howard (National Institutes of Health). This plasmid was transfected using LipofectAMINE (Life Technologies, Inc.) at 1 μg of DNA per ml and 10 μl of liposomes per ml in Opti-MEM (Life Technologies, Inc.) for 5 h in the absence of serum. The Chimera-1 DNA was constructed by ligating the DNA encoding the extracellular portion of the FcεRI α subunit, the DNA encoding the transmembrane and intracellular portions of the type I IL-1 receptor, and the pcDNA3.1 mammalian expression vector (Invitrogen, Carlsbad, CA). The rat FcεRIα cDNA used to make this construct was provided by Dr. H. Metzger (National Institutes of Health). The type I IL-1 receptor DNA was cloned using reverse transcription PCR from murine adult brain total RNA (provided by Dr. A. Trumpp, University of California, San Francisco). Sequencing confirmed that the junction of this chimeric receptor corresponds to base 669 of the FcεRI α sequence (GenBankTM accession number M17153) and base 1015 of the IL-1 receptor sequence (GenBankTM accession numberM20658). The Chimera-1 plasmid was transfected using LipofectAMINE Plus (Life Technologies) at 2 μg of DNA per ml and 3 μl of liposomes per ml in Opti-MEM for 3 h. Primary and secondary antibodies used to form receptor complexes at the cell surface were as follows. For IgE, affinity-purified rabbit anti-mouse IgE (34Menon A.K. Holowka D. Baird B. J. Cell Biol. 1984; 98: 577-583Crossref PubMed Scopus (42) Google Scholar); for α4 integrin, TA-2 mouse monoclonal antibody (35Issekutz T.B. Wykretowicz A. J. Immunol. 1991; 147: 109-116PubMed Google Scholar) provided by Dr. T. Issekutz (Toronto Hospital, Toronto, Ontario, Canada) and rabbit anti-mouse IgG Fc secondary antibody (Jackson ImmunoResearch, West Grove, PA); for IL-1 receptors, M5 rat monoclonal antibody (36Gallis B. Prickett K.S. Jackson J. Slack J. Schooley K. Sims J.E. Dower S.K. J. Immunol. 1989; 143: 3235-3240PubMed Google Scholar), provided by Dr. S. Dower, and goat anti-rat IgG secondary antibody (Fisher); for IL-2 receptor α, 3G10 mouse monoclonal antibody (Boehringer Mannheim) and rabbit anti-mouse IgG secondary antibody (Cappel, West Chester, PA). All primary antibodies were iodinated with chloramine T as described previously (37Holowka D. Baird B. Biochemistry. 1983; 22: 3466-3474Crossref PubMed Scopus (84) Google Scholar). After cell harvest, the receptors were labeled with the appropriate primary antibody for at least 30 min at 20 °C, followed by two washes in buffered salt solution (20 mm Hepes, pH 7.4, 135 mm NaCl, 5 mm KCl, 1.8 mmCaCl2, 1 mm MgCl2, 5.6 mm glucose). For α4 integrin experiments, RBL cells were presaturated with IgE, and IgE was present during TA-2 binding to prevent any interaction of FcεRI with the Fc portion of TA-2. Confluent cells were harvested using EDTA and suspended at 8 × 106/ml in buffered saline solution with 1 mg/ml bovine serum albumin. Cells labeled with the appropriate antibody were stimulated with streptavidin (Sigma) or secondary antibodies as indicated under "Results," followed by the addition of an equal volume of ice cold 2× lysis buffer (final concentration, 25 mm Hepes, pH 7.5, 50 mm NaCl, 10 mm EDTA, 1 mmNa3VO4, 30 mm pyrophosphate, 10 mm glycerophosphate, 1 mm4-(2-aminoethyl)benzenesulfonyl fluoride (Calbiochem), 0.02 units/ml aprotinin, and 0.01% (w/v) NaN3) with the indicated concentration of Surfact-Amps TX-100 (Pierce). For aggregating FcεRI after cell lysis, we used rabbit anti-IgE because lysed RBL cells appear to contain sufficient free biotin to interfere with streptavidin aggregation. 3K. A. Field, D. Holowka, and B. Baird, unpublished observations. After incubation on ice for at least 10 min, the lysates were then diluted with an equal volume of 80% (w/v) sucrose in 25 mmHepes, pH 7.5, and 150 mm NaCl. Step gradients of sucrose were formed by layering 0.25 ml of 80%, 0.5 ml of 60%, 1.5 ml of 40% (containing the cell lysate), 0.75 ml of 30%, 0.5 ml of 20%, and 0.5 ml of 10% (w/v) sucrose in Beckman Ultra-Clear centrifuge tubes (11 × 60 mm). Centrifugation and the analysis 0.2-ml aliquots of the gradients were performed as described (6Field K.A. Holowka D. Baird B. J. Biol. Chem. 1997; 272: 4276-4280Abstract Full Text Full Text PDF PubMed Scopus (325) Google Scholar, 16Field K.A. Holowka D. Baird B. Proc. Natl. Acad. Sci. U. S. A. 1995; 92: 9201-9205Crossref PubMed Scopus (271) Google Scholar). Sucrose gradient fractions containing DRMs from either anti-IgE-stimulated or unstimulated cells lysed in 0.05% TX-100 were pooled according to the distribution of125I-IgE bound to FcεRI in a parallel stimulated sample. Subsequent steps were performed with the technical assistance of Shannon Caldwell at the Cornell Integrated Microscopy Center. Formvar carbon-coated grids (300 mesh) were suspended on the top of drops of the sucrose fractions for 30 min to allow the adherence of DRMs. The grids were then extensively washed, fixed with 1% gluteraldehyde for 3 min, washed, and stained for 30 s with 2% uranyl acetate. The negatively stained DRMs were visualized with a Philips EM-201 transmission electron microscope. We previously demonstrated that the PTK Lyn associates with DRMs isolated from RBL cells, and we found that this association is enhanced upon FcεRI stimulation (16Field K.A. Holowka D. Baird B. Proc. Natl. Acad. Sci. U. S. A. 1995; 92: 9201-9205Crossref PubMed Scopus (271) Google Scholar). Under these standard cell lysis conditions with high concentrations of TX-100, FcεRI did not co-isolate with DRMs. However, with lower detergent lysis conditions, similar to those identified by Pribluda et al. (7Pribluda V.S. Pribluda C. Metzger H. Proc. Natl. Acad. Sci. U. S. A. 1994; 91: 11246-11250Crossref PubMed Scopus (173) Google Scholar) for enhancing the association of kinase activity with immunoprecipitated FcεRI, we found that substantial amounts of the aggregated receptor co-purified with the low density DRMs following sucrose gradient ultracentrifugation (6Field K.A. Holowka D. Baird B. J. Biol. Chem. 1997; 272: 4276-4280Abstract Full Text Full Text PDF PubMed Scopus (325) Google Scholar). As part of our investigation of the structural basis for this receptor association with DRMs, we examined the detergent sensitivity. For low detergent conditions, we use 0.05% TX-100 to lyse RBL cells at 4 × 106/ml and then dilute this lysate 1:1 with 80% sucrose prior to sucrose gradient ultracentrifugation. Under these conditions, the micellar detergent to cell lipid ratio during the ultracentrifugation is approximately 3 (38Kinet J.P. Alcaraz G. Leonard A. Wank S. Metzger H. Biochemistry. 1985; 24: 4117-4124Crossref PubMed Scopus (38) Google Scholar). To exclude some artifactual possibilities, we added 0.025% TX-100 to all of the sucrose gradient steps, as shown in Fig.1 (○ and •). Under these conditions, as when no TX-100 is added to the gradient (6Field K.A. Holowka D. Baird B. J. Biol. Chem. 1997; 272: 4276-4280Abstract Full Text Full Text PDF PubMed Scopus (325) Google Scholar), less than 5% of the unaggregated FcεRI associates with the DRM fractions (○, fractions 3–7) and greater than 50% of streptavidin-aggregated biotin-IgE-FcεRI complexes associate with these (•). From this result, it appears unlikely that the plasma membrane is incompletely solubilized or that association of aggregated receptors with DRMs depends on the separation of DRM components from TX-100 during the ultracentrifugation process. To investigate the requirements for DRM association of aggregated FcεRI, we lysed cells with 0.05% TX-100 prior to adding anti-IgE to form FcεRI aggregates. When this lysate was fractionated (Fig. 1, ■), these aggregates migrated at a higher density in the sucrose gradient, similar to aggregates formed on cells that are subsequently lysed in 0.2% or higher TX-100 (16Field K.A. Holowka D. Baird B. Proc. Natl. Acad. Sci. U. S. A. 1995; 92: 9201-9205Crossref PubMed Scopus (271) Google Scholar). In contrast, addition of anti-IgEprior to cell lysis with 0.05% TX-100 causes greater than 50% of FcεRI to associate with low density membranes (e.g. see Fig. 6 A), similar to results using streptavidin cross-linking (Fig. 1 (•) and Ref. 6Field K.A. Holowka D. Baird B. J. Biol. Chem. 1997; 272: 4276-4280Abstract Full Text Full Text PDF PubMed Scopus (325) Google Scholar). These results indicate that the interactions between DRM components and aggregated FcεRI that occur on intact cells must be preserved during cell lysis and sucrose gradient fractionation, as they cannot be caused by receptor aggregation subsequent to cell lysis. The association of FcεRI with DRMs could be mediated by protein-protein, protein-lipid, or some combination of these interactions. By varying the amount of TX-100 used to lyse RBL cells, we further investigated the detergent sensitivity of this interaction. Unaggregated FcεRI was recovered in the low density gradient fractions when insufficient TX-100 (<0.03%) was used for cell lysis (Fig. 2, ○). Under these conditions, it is possible that plasma membranes are not solubilized completely and therefore migrate at this low density because of their lipid content. With greater than 0.03% TX-100 for cell lysis, the membranes appeared effectively solubilized, and monomeric FcεRIs were found almost entirely in the 40% sucrose fraction. In addition, other membrane-bound proteins that do not associate with DRMs, including α4 integrins (e.g. see Fig. 6 B) and Src,3 were found in the 40% sucrose fractions at 0.05% TX-100. Under these same conditions, greater than 50% of streptavidin-aggregated biotin-IgE bound to FcεRI remained associated with DRMs in the sucrose gradient (Fig. 2, •). When concentrations of TX-100 used during cell lysis were greater than 0.05%, aggregated FcεRIs were not retained with DRMs, and complete disruption of this association occurred as low as 0.08% TX-100 (Fig. 2). As shown previously, lipid-anchored markers of these DRMs (i.e.Thy-1, GD1b gangliosides, and Lyn) remain associated even at 0.5% TX-100 (16Field K.A. Holowka D. Baird B. Proc. Natl. Acad. Sci. U. S. A. 1995; 92: 9201-9205Crossref PubMed Scopus (271) Google Scholar). This marked sensitivity for FcεRI/DRM association is consistent with a lipid-mediated interaction (see under "Discussion"). Electron microscopy of isolated, negatively stained DRMs from RBL cells lysed at 0.05% TX-100 reveals them to be vesicular structures (Fig.3). The majority range from 50 to 200 nm in diameter; larger vesicles with diameters up to 1.4 μm were less frequently observed.3 These vesicles are qualitatively similar in appearance and size to DRM preparations from MDCK cells (25Lisanti M.P. Scherer P.E. Vidugiriene J. Tang Z. Hermanowski-Vosatka A. Tu Y.H. Cook R.F. Sargiacomo M. J. Cell Biol. 1994; 126: 111-126Crossref PubMed Scopus (813) Google Scholar), T- and B-cell lines (28Parolini I. Sargiacomo M. Lisanti M.P. Peschle C. Blood. 1996; 87: 3783-3794Crossref PubMed Google Scholar), neuroblastoma cells (26Gorodinsky A. Harris D.A. J. Cell Biol. 1995; 129: 619-627Crossref PubMed Scopus (296) Google Scholar), and RBL cells, more typically lysed in 1% TX-100. 4D. Reczeck and K. Field, unpublished observations. Thus, DRMs isolated with lower TX-100 to cell lipid ratios are not ultrastructurally different from other DRMs that have been prepared with higher TX-100, and these preparations are not significantly contaminated with unsolubilized membrane sheets or organelles. To investigate further the structural basis of the FcεRI/DRM interaction, we used P815 mast cells and CHO cells that were stably transfected with various FcεRI constructs. As shown in Fig.4, P815 cells stably expressing wild-type αβγ2 FcεRI subunits (wt) show aggregation-dependent receptor association with low density DRMs that is similar to native FcεRI on RBL cells at 0.05% TX-100 (Fig. 2 and Ref. 6Field K.A. Holowka D. Baird B. J. Biol. Chem. 1997; 272: 4276-4280Abstract Full Text Full Text PDF PubMed Scopus (325) Google Scholar). Furthermore, cells expressing mutated versions of the receptor lacking either the C-terminal cytoplasmic tail of the β subunit or the cytoplasmic tail of the γ subunit show similar aggregation-dependent association (Fig. 4). Interestingly, IgE receptors entirely lacking the β subunit (FcεRI αγ2) also show aggregation-dependent DRM interactions. Some differences in the magnitude of association were observed, but the ratio of aggregated to monomeric receptors associated with the DRMs remained fairly constant. These differences could represent small contributions to the DRM association from the deleted protein segments. Alternatively, this variability could be due to differences between the transfected cell lines themselves that have been separately subcloned. Significantly, FcεRI in the β and γ subunit cytoplasmic tail mutant cells do not activate tyrosine kinases or other downstream signals (9Alber G. Miller L. Jelsema C.L. Varin-Blank N. Metzger H. J. Biol. Chem. 1991; 266: 22613-22620Abstract Full Text PDF PubMed Google Scholar), although they do become insoluble in high concentrations of TX-100 after extensive aggregation (32Mao S.Y. Alber G. Rivera J. Kochan J. Metzger H. Proc. Natl. Acad. Sci. U. S. A. 1992; 89: 222-226Crossref PubMed Scopus (27) Google Scholar). Nonhematopoietic CHO cells expressing wild-type FcεRI showed qualitatively similar association of this receptor with DRMs that depends on aggregation (Fig. 5,wild-type). However, a chimeric receptor containing the extracellular segment of the FcεRI α subunit and the transmembrane and intracellular segments of the type-1 IL-1 receptor (Fig. 5,Chimera-1) did not show significant DRM association in the presence or absence of receptor aggregation. This chimeric receptor demonstrates that the extracellular segment of FcεRI α is not sufficient to mediate this interaction. In contrast, a similar chimeric receptor that contains the extracellular segment of the FcεRI α subunit and the transmembrane and intracellular segments of IL-2 receptor α (p55, Tac; 39) (Fig. 5, Chimera-2) did show aggregation-dependent association with the DRMs. The results in Figs. 4 and 5 are consistent with the transmembrane segments of these receptors being most important for determining DRM association. We also examined the association of FcεRI α subunits linked to the plasma membrane by a GPI anchor. These receptors, which lack a transmembrane segment, associated with the DRMs even in the absence of aggregation (Fig. 5, GPI) and also remained associated to a similar extent in 0.2% TX-100,3 conditions that would extract aggregated, wild-type FcεRI (Fig. 2). This is consistent with previous studies from our laboratory (16Field K.A. Holowka D. B
Referência(s)