Lateral Electromigration and Diffusion of Fcc Rat Basophilic Leukemia Cells: Effects of IgE Receptors on Binding

We have used in situ electromigration and post-field relaxation (Poo, M.-m. 1981, Annu. Rev. Biophys. Bioeng., 10:245-276) to assess the effect of immunoglobulin E (IgE) binding on the lateral mobility of IgE-Fc receptors in the plasmalemma of rat basophilic leukemia (RBL) cells. Bound IgE sharply increased the receptor's electrokinetic mobility, whereas removal of cell surface neuraminic acids cut it to near zero. In contrast, we found only a small difference between the lateral diffusion coefficients (D) of vacant and IgEoccupied Fc receptors (D: 4 vs. 3 x 10 -1° cm2/s at 24°C). This is true for monomeric rat IgE; with mouse IgE, the difference in apparent diffusion rates was slightly greater (D: 4.5 vs. 2.3 x 10 -t° cm2/s at 24°C). This range of D values is close to that found in previous photobleaching studies of the IgE-Fc~ receptor complex in RBL cells and rat mast cells. Moreover, enzymatic depletion of cell coat components did not measurably alter the diffusion rate of IgE-occupied receptors. Thus, binding of fluorescent macromolecular probes to cell surface proteins need not severely impede lateral diffusion of the probed species. If the glycocalyx of RBL cells does limit lateral diffusion of the Fc~ receptor, it must act primarily on the receptor itself, rather than on receptor-bound IgE. Integral proteins diffuse 10-10 3 times more slowly in plasma membranes than in reconstructed lipid bilayers. Hydrodynamic theory also predicts that integral proteins should diffuse much faster than they do in vivo, provided the major drag that they experience is that due to lipid viscosity (1). At least this is the indication from a host of diffusion studies employing the method of fluorescence recovery after photobleaching (FRAP).' Hence, the search is on for extramembranous structures that impede diffusion of membrane components, with much attention being focused on the cytoskeletal matrix. In this paper, we ask whether or not the "anomalously" slow diffusion observed in natural membranes is an artifact of the FRAP technique, rather than a true reflection of the mobilities of membrane proteins. Several experiments (summarized in reference 2) have essentially silenced most critics who previously felt that protein Abbreviations used in this paper: CBA, cell buffer with azide; Con A, concanavalin A: D, diffusion coefficient; FRAP, fluorescence recovery after photobleaching; PFR, post-field relaxation; RBL, rat basophilic leukemia; TRITC, tetramethylrhodamine isothiocyanate; TRr-IgE, Texas Red-labeled rat IgE. 778 cross-linkage resulting from the intense bleaching light might explain the slow diffusion reported in FRAP studies. A still untested hypothesis, however, is that high molecular weight fluorescent ligands normally used in FRAP impede the intrinsic mobility of the membrane proteins to which they are bound. This does not occur for antibodies bound to lipid haptens in pure lipid bilayers (3), but might happen ifa bound ligand were to increase the drag on its receptor by interaction with polymeric constituents of the cell coat, or glycocalyx. Notably, addition of an artificial "cell coat" to the antibodylipid hapten system markedly reduces the observed lateral diffusion coefficient (D) value of lipid-bound antibodies (4). A few previous studies lend support to the suggestion that nonligated membrane receptors may diffuse much faster than ligand-receptor complexes. Using the techniques of post-field relaxation (PFR) (5) and post-inactivation recovery (6), the D values of ligand-free concanavalin A (Con A) and acetylcholine (ACh) receptors in embryonic Xenopus muscle cells were found to lie in the range of 1-4 x 10 -9 cm2/s (at room temperature), much faster than the FRAP-determined values for acetylcholine receptors in embryonic rat muscle (7) or Con A receptors in the same Xenopus muscle (Liu, Z.-y., and THE JOURNAL OF CELL 81OLOQY . VOLUME 99 SEPTEMBER 1984 778-787 © The Rockefeller University Press . 0021-9525/84/09/0778/10 $1.00 on O cber 4, 2017 jcb.rress.org D ow nladed fom M-m. Poo, unpublished results). Using PFR, Tank (8) found that uncomplexed low-density lipoprotein receptors on an internalization-defective human fibroblast cell line diffuse at 1.1 x 10 -9 cm2/s; but using FRAP a D value of 1.4 x 10 -~ cm2/s was observed for the low-density lipoprotein receptor complex (9). The goal of our present work was to use one technique, namely, PFR, to compare the diffusion coefficients of one receptor in ligand-free and ligand-bound states. The choice of IgE-Fc receptors on rat basophilic leukemia (RBL) cells offers two advantages in this regard. First, the lateral diffusion rates of IgE-Fc receptor complexes on RBL cells (and rat mast cells) have already been measured with FRAP (2, 10), so one can compare the results from two different methods for one receptor on one cell type. Second, the interaction of IgE with its receptor is strong yet univalent. By using multimer-free IgE one can therefore maintain a high molecular weight (185,000) ligand for long periods on the cell surface without inducing receptor cross-linkage or the active cellular responses of anchorage modulation and internalization that can attend the binding of lectins, F(ab')2 fragments, or other multivalent ligands. The present study therefore provides a direct test of the idea that nonspecific interaction of a macromolecular label (IgE) with the cell coat retards lateral diffusion of the IgE-bound Fc receptor. MATERIALS AND METHODS Cells: The 2H3 sublin¢ of RBL-IV cells (1 l) was provided by Dr. Henry Metzger, National Institutes of Health. Cell monolayers were grown and passaged as described in (12). Stock cultures were passaged weekly into 75-cm 2 flasks (Coming Glass Works, Coming, NY), and at this time aliquots of 1.5 x l06 cells were seeded into 6-cm-diam tissue culture dishes (Coming), using the same medium as for the stock culture. In pilot experiments, we observed a systematic increase in the starting asymmetry index (see below) with culture age; in the diffusion studies we therefore only used cultures 3-4 d old. The medium was decanted, and the monolayer was rinsed twice in liquid policeman (0.6 mM Na2 EDTA, 0.14 M NaCl, 2.7 mM KCI, 9.5 mM NaKP~, pH 7.2), and incubated for 12 rain at 37°C in 4 ml of the same buffer. Cells were dislodged with a gentle stream from a pasteur pipette, diluted to 25-50 ml with Tyrode's buffer (0.137 M NaCl, 2.7 mM KCI, 1.5 mM CaCl2, 0.4 mM P~, l mM MgCl2, 5.6 mM glucose, 0.2% BSA, 0.02 M HEPES, pH 7.4) at room temperature, and harvested by centrifugation for 4-5 min at 200 g in polypropylene tubes. For pre-field labeling experiments cell pellets were resuspended in Tyrode's plus 0.2% BSA (Sigma Chemical Co., St. Louis, MO), and for post-field labeling in cell buffer with azide (CBA) (10 mM HEPES, 28 mM NaNs, 1 mM NaPi, 109 mM NaCl, 2 mM CaCl2, 1.5 mM MgCl2, 5.4 mM KCI, 0.1% glucose, pH 7.3). Cell viability was checked with trypan blue, and cells were diluted to 1-2 x 106/ml and then either plated directly in electromigration chambers (postfield labeling) or labeled in suspension with fluorescent lgE (pre-field labeling). In the former case cells were used within 2 h of isolation, and most data is for those used within 30 min. In pre-field labeling, cells were kept on ice for up to 5 h before use. Fluorescent Antibodies: Purified rat IgE from the IR162 immunocytoma (13) was supplied by Dr. Henry Metzger. For later experiments we purified IR162 IgE from ascites fluid also supplied to us by Dr. Metzger. Affinity-purified mouse monoclonal anti-2,4-dinitrophenyl IgE from the HI dinitrophenyl-~-26.82 hybridoma (14) was provided by Dr. Fu-tong Liu of the Medical Biology Research Institute, La Jolla, CA. IgE was conjugated with either tetramethylrhodamine isothiocyanate (TRITC, isomer R, lot A 1GCZK; Baltimore Biological Laboratories, Cockeysville, MD) or Texas Red (Molecular Probes, Junction City, OR). Fluorescein isothiocyanate-labeled rabbit antimouse IgG was obtained from Miles Laboratories (Elkhart, IN). For TRITC conjugation, to l mg of mouse IgE in l ml PBS was added 2 drops of 1 M carbonate buffer, pH 9. l0 #l of a 20 mg/ml solution of the isothiocyanate in dimcthylsulfoxide was added, the mixture homogenized by inversion and left on ice for 4 h. Free dye was removed by passage through a P6 column (Bio-Rad Laboratories, Budingame, CA) equilibrated with PBS and dialysis for 24 h (4°C) against PBS and NAN3. Using the absorbance at 280, 515, and 555 nm 05) the molar ratio of dye/protein was estimated to be ca. 2 and 5 for two different preparations. Subsequent gel filtration produced no change in these numbers, indicating that noncovalently bound dye was not bleeding off the conjugate. Texas Red was coupled with either rat or mouse IgE by mixing an equal volume of borate buffered saline (pH 8.0) containing 4-5 mg/ml of IgE with 0.25 M Na carbonate buffer, pH 9, and then adding powdered Texas Red (0.25 mg/mg IgE). The suspension of dye was inverted periodically during a l-h (rat) or 2-h (mouse) incubation on ice. Free dye was removed from the blue conjugate on a P6 column as above. From absorbance at 280 and 596 rim, the estimated dye/protein ratios were 2.5 for rat and 5.5 for mouse IgE (16). Before labeling cells, we routinely centrifuged the lgE for 30-60 min at 100,000 g (4°C) in a Beckman alrfuge (Beckman Instruments, Inc., Fullerton, CA), and then used only the upper 60-70% of the supemate. In some experiments we used a Texas Red-rat IgE (TRr-lgE) conjugate which had been purified (Fig. 1) by chromatography on sequential columns of AcA 34 and AcA 22 (LKB Produkter, Bromma, Sweden) (17). The pooled monomer fractions were used directly (0.6 mg IgE/ml), but even after reconcentration to 4 mg/ml, analysis by nondenaturing PAGE (4-25%) showed no evidence of aggregated IgE. Since the D values obtained were equal to those for less rigorously purified TRr-IgE, we assume that aggregation of the latter was not a problem. Neither TRITCnor Texas Red-mouse IgE was subjected to PAGE, but PAGE of the underivatized mouse IgE did reveal multimeric components. Because the mouse IgE was not purified by gel filtration after its reaction with TRITC or Texas