Salt bridge residues between I-Ak dimer of dimers alpha-chains modulate antigen presentation

Class II dimers of dimers are predicted to have functional significance in antigen presentation. The putative contact amino acids of the I-Ak class II dimer of dimers have been identified by molecular modeling based on the DR1 crystal structure (Nydam et al., Int. Immunol. 10, 1237,1998). We have previously reported the role in antigen presentation of dimer of dimers contact amino acids located in the C-terminal domains of the alpha- and beta-chains of class II. Our calculations show that residues Ealpha89 and Ralpha145 in the alpha2-domain form an inter alpha-chain salt bridge between pairs of alphabeta-heterodimers. Other residues, Qalpha92 and Nalpha115, may be involved in close association in that part of the alpha-chain. We investigated the role of these amino acids on class II expression and antigen presentation. Class II composed of an Ealpha89K substituted alpha-chain paired with a wt beta-chain exhibited inhibited antigen presentation and expression of alpha-chain serologic epitopes. In contrast, mutation of Ralpha145E had less affect on antigen presentation and did not affect I-Ak serologic epitopes. Interchanging charges of the salt bridge residues by expressing both Ralpha145E and Ealpha89K on the same chain obviated the large negative effect of the Ealpha89K mutation on antigen presentation but not on the serologic epitopes. Our results are similar for those reported for mutation of DR3's inter-chain salt bridge with the exception that double mutants did not moderate the DR3 defect. Interestingly, the amino acids differences between I-A and DR change the location of the inter-chain salt bridges. In DR1 these residues are located at positions Ealpha88 and Kalpha111; in I-Ak these residues are located at position Ealpha89 and Ralpha145. Inter alpha-chain salt bridges are thus maintained in various class II molecules by amino acids located in different parts of the alpha2-domain. This conservation of structure suggests that considerable functional importance may attach to the ionic interactions.

Intrinsically fluorescent luteinizing hormone receptor demonstrates hormone-driven aggregation

The possibility that LH receptors exist as isolated molecules when unbound and aggregate upon binding gonadotropins has previously been untestable in viable cells for want of a suitable nonhormone probe. We have now expressed in CHO cells an intrinsically-fluorescent LH receptor involving enhanced green fluorescent protein (GFP) fused to the C-terminus of the rat LH receptor (rLHR-GFP). More than half of these receptors (54 +/- 4%) are located on the plasma membrane and are functional: cAMP levels increase 3-5 fold in response to 10 nM LH or hCG. In fluorescence photobleaching recovery studies at 37 degrees C, 54 +/- 13% of unoccupied rLHR-GFP were laterally mobile with a diffusion coefficient D of 16 +/- 3.5 x 10(-10)cm2sec-1. Introduction of 10 nM LH for 1 h slowed receptor lateral diffusion to 6.6 +/- 1.3 x 10(-10)cm2sec-1 and reduced fluorescence recovery after photobleaching to 27 +/- 1%. Following treatment with 1 nM hCG, rLHR-GFP were laterally immobile and were distributed into small fluorescent patches over the cell surface. Thus, unoccupied rLHR-GFP receptors apparently exist as dispersed plasma membrane proteins with comparatively fast lateral diffusion. Interaction of receptors with LH or hCG caused clustering of rLHR-GFP receptors, significantly restricting lateral diffusion.

Characterization of an intrinsically fluorescent gonadotropin-releasing hormone receptor and effects of ligand binding on receptor lateral diffusion

The GnRH receptor (GnRHR) is a G protein-coupled receptor expressed by gonadotropes in the anterior pituitary gland. In the past several years, much has been learned about the structure-function relationships that exist in this receptor with regard to ligand binding and signal transduction. However, the lack of specific antibodies has precluded any analyses of the behavior of the unbound form of this receptor. We have constructed a functional GnRHR in which enhanced green fluorescent protein is fused to the carboxyl-terminus of the murine GnRHR. This fusion receptor was expressed diffusely throughout the cell, with approximately 38% of the fusion receptors colocalized with a plasma membrane marker in the gonadotrope-derived alphaT3 cell line, and approximately 82% of the fusion receptors colocalized with a membrane marker in Chinese hamster ovary cells. Furthermore, the fusion receptor displayed a Kd of 0.8 nM for iodinated des-Gly10,D-Ala-6-GnRH N-ethyl amide in Chinese hamster ovary cells, which was similar to the Kd of the native GnRHR expressed in alphaT3 cells. The surface mobility of the fusion protein was examined by fluorescence photobleaching recovery methods. In the unbound state the majority of the receptors were laterally mobile and displayed a lateral diffusion rate of 1.2-1.6 x 10(-9) cm2/sec. Binding of GnRH reduced the rate of lateral diffusion over 3-fold and reduced the fraction of mobile receptors from approximately 76-91% to 44-61%. Like GnRH, the competitive GnRH antagonist antide slowed the rate of receptor diffusion approximately 3-fold. In contrast to GnRH, antide had no effect on the fraction of mobile receptors. Thus, an intrinsically fluorescent GnRHR is trafficked to the plasma membrane of mammalian cells, is capable of ligand binding and signal transduction, and allows direct observation of the GnRHR in the nonligand-bound state. Furthermore, fluorescence photobleaching recovery analysis of the GnRHR-green fluorescent protein fusion reveals fundamental differences in the membrane dynamics of the GnRHR induced by the binding of an agonist vs. that induced by the binding of an antagonist.

Mutations in MHC class II dimer of dimers contact residues: effects on antigen presentation

The recent solutions of the MHC class II crystal structure reveal dimerization of the alphabeta heterodimers. These dimer of dimers structures may also exist either on resting cells or after engagement by TCR, and may be involved in B cell signaling and up-regulation of co-stimulatory molecules such as B7 which facilitate T cell activation. By combining crystallographic data on HLA-DR1 with the sequence of murine I-Ak and refining the resulting structure through energy minimization calculations, we have predicted the contact amino acids expected to stabilize the I-Ak dimer of dimers structure. As in HLA-DR1, three salt bridges in I-Ak (D alpha62-Hbeta112, H alpha181-E beta163, E alpha183-Hbeta113) appear to provide the main interaction. Guided by this structural data, we prepared 45 B cell transfectants representing 20 different class II mutation phenotypes in the contact region containing these salt bridges. We examined their abilities to activate three T cell hybrids. Antigen-specific h4Ly50.5 cells were not greatly affected by changes in the dimer of dimer contact residues. In contrast, autoreactive C8.A3 T cells were very sensitive to changes in this region but presentation of class II of many mutation phenotypes could be rescued by treatments that up-regulate B7-1. The alloreactive hybridoma 2H40.2.5 was less sensitive to changes in the contact residues. A simple model was developed that summarizes the effects of the mutations for the T cells tested. Mutations at D alpha162, E alpha183, H alpha181 and Rbeta106 had the largest negative impact, while D alpha166, E alpha185, Hbeta112, Hbeta113 and E beta163 were less disruptive. Results are consistent with mutations interfering with class II interaction with another molecule which might or might not be another class II heterodimer. However, the larger negative impact of alpha chain mutations in salt bridge pairs suggests that these sites also help maintain some essential conformation of the alpha chain apart from any possible impact on dimer of dimers stability.

Interferometric fringe fluorescence photobleaching recovery interrogates entire cell surfaces

Fluorescence photobleaching recovery (FPR) measurements of cell surface protein lateral diffusion typically employ an interrogated spot of 0.5 microm 1/e2 radius. The effective spot area represents only 1/500 of the total surface of an 8-microm cell. An FPR measurement of a protein expressed as 50,000 copies per cell reflects the dynamics of 100 molecules. This limits the precision and reproducibility of FPR measurements. We describe a method for interferometric fringe pattern FPR that permits simultaneous interrogation of the entire cell's surface. Fringe patterns are generated interferometrically within the optical path of an FPR system. Methods for interpreting fluorescence recovery kinetics on cells and for determining the protein mobile fraction are presented. With fringe FPR, the murine major histocompatibility complex class II antigen I-Ak expressed on M12.C3.F6 cells has 100-fold improved fluorescence signals relative to spot FPR, with corresponding improvements in signal-to-noise ratios of recovery traces. Diffusion coefficients (+/- standard deviation) of (2.1 +/- 0.4) x 10(-10) and (1.8 +/- 1.0) x 10(-10) cm2 s-1 with corresponding mobile fractions of I-Ak of 66.1 +/- 7.8% and 63.4 +/- 18.0% were obtained by fringe and spot methods, respectively. The improved reproducibility of fringe over spot results is less than signal improvements predict. There may thus be substantial variation from cell to cell in protein dynamics, and this method may permit the assessment of such variation.

Luteinizing hormone receptors are associated with non-receptor plasma membrane proteins on bovine luteal cell membranes

Biophysical studies of the bovine luteinizing hormone (LH) receptor on luteal cell membranes suggest that this receptor may be part of a larger molecular weight structure. We have used 5-iodonaphthyl-1-azide (INA) to identify plasma membrane proteins near LH receptors on plasma membranes from bovine corpora lutea. Following binding of eosin isothiocyanate-derivatized ovine LH or human chorionic gonadotropin (hCG), five proteins with molecular weights of 71, 57, 55, 49 and 36 kDa were selectively derivatized with [125I]-INA following 2 h exposure at 22 degreesC to 514 nm light. However, there was no fluorescence energy transfer between LH receptors occupied by ovine LH or hCG indicating that LH receptors were not self-associated in these membrane preparations. Together these results suggest that, following hormone binding, single copies of the LH receptor may exist in large molecular weight structures that include non-receptor proteins.

The rotational diffusion of LH receptors differs when receptors are occupied by hCG versus LH and is increased by cytochalasin D

We have examined the rotational diffusion of the luteinizing hormone (LH) receptors binding human chorionic gonadotropin (hCG) or ovine luteinizing hormone (oLH) in MA-10 Leydig tumor cells using time-resolved phosphorescence anisotropy techniques. LH receptors binding erythrosin isothiocyanate (ErITC)-derivatized oLH were rotationally mobile with rotational correlation times of 62 micros, 48 micros, 38 micros, and 29 micros at 4 degrees C, 15 degrees C, 25 degrees C, and 37 degrees C, respectively. ErITC-hCG bound to the LH receptor was rotationally immobile, showing no anisotropy decay at 4 degrees C, 15 degrees C, 25 degrees C, and 37 degrees C. To determine whether cytoskeletal components influenced the rotational diffusion of LH receptors, we measured rotational diffusion of LH receptors on MA-10 cells treated with 20 microg/ml cytochalasin D and on plasma membrane preparations. Following 1 h exposure to cytochalasin D, the rotational correlation times for hCG-occupied LH receptors were typically 11 micros at 37 degrees C compared to > 1000 micros on untreated cells. Treatment of MA-10 cells with cytochalasin B or colchicine had no affect on LH receptor rotational diffusion. Rotational correlation times for LH-occupied receptors decreased from 29 micros to 12 micros at 37 degrees C following cytochalasin D treatment. The rotational diffusion of LH receptors on plasma membrane preparations was similar to that observed for LH- and hCG-occupied receptors on intact cells treated with cytochalasin D. These various results indicate that there are differential effects of LH and hCG binding on the interactions of LH receptors with plasma membrane proteins and that microfilaments anchor the hCG- and LH-occupied receptors.

5-iodonaphthyl-1-azide labeling of plasma membrane proteins adjacent to specific sites via energy transfer

We have examined conditions optimal for 5-iodonaphthyl-1-azide (INA4) labeling of membrane proteins proximal to known membrane sites. Membrane-bound INA can be indirectly activated by energy transfer from visible chromophores. We demonstrate that the efficiency of this sensitized activation is enhanced by use of triplet-forming chromophores such as eosin and by deoxygenation. Variation of sensitized activation efficiency with INA concentration indicates that the critical distance for eosin-INA energy transfer in solution is 8-14 A. We suggest that photosensitization occurs through triplet exchange and present an improved labeling protocol based on these findings. This protocol was used to examine whether different accessory proteins are associated with isolated and crosslinked Type I Fc epsilon receptors on 2H3 rat basophilic leukemia cells. 2H3 cells were incubated with eosin-conjugated IgE and irradiated at 514 nm yielding [125I]INA derivatized peptides at 53, 38, 34, and 29 kDa. Crosslinking IgE with mouse anti-rat IgE prior to irradiation labeled three additional proteins at 60, 54, and 43 kDa. These results demonstrate the utility of sensitized INA labeling in characterizing protein-protein interactions in membranes of intact cells and indicate the importance of considering photophysical factors when selecting sensitizers and reaction conditions. We discuss estimation of the size of the membrane region surrounding a sensitizing chromophore within which INA labeling of membrane proteins occurs.

Lateral dynamics of major histocompatibility complex class II molecules bound with agonist peptide or altered peptide ligands

We examined the lateral diffusion of I-Ad on A20 cells following the binding of ovalbumin-derived peptides. The peptides were OVA323-339 and OVA325-335 and a related peptide OVA325-335s substituted H331Q. Only OVA323-339 and OVA325-335 were effectively presented by A20 cells to DO-11.10/S4.4 T cells as assessed by IL-2 production. Fluorescence photobleaching recovery (FPR) measurements showed anti-I-Ad to have a lateral diffusion coefficient on untreated A20 cells of 1.8 +/- 1.0 x 10(-10) cm2 s-1 at 25 degrees C with fluorescence recovery after photobleaching greater than 50%. After 24 h incubation of A20 cells with OVA323-339 or OVA325-335, a subpopulation of A20 cells appeared that were approximately half the size of untreated A20 cells. Culture of A20 with OVA325-355s did not stimulate DO-11.10 cells or induce a size change in A20 cells. Class II molecules were laterally immobile on these small cells with fluorescence recoveries after photobleaching of less than 20%. The relative number of small cells in the A20 cell population was correlated with the immunogenicity of the peptides. These results suggest that immobilization of surface I-Ad may be an important event in antigen presentation.

Rotational dynamics of luteinizing hormone receptors on bovine and ovine luteal cell plasma membranes

To determine whether LH receptor rotational diffusion is similar in closely related species, we compared the rotational correlation times of LH receptors on bovine CL membranes with those of LH receptors on sheep small luteal cells and luteal cell plasma membranes using time-resolved phosphorescence anisotropy techniques. After binding of erythrosin isothiocyanate (ErITC)-derived bovine LH (bLH), ErITC-ovine LH (oLH), or ErITC-hCG, there was no difference in the initial and final anisotropy at 4 degrees C, 15 degrees C, 25 degrees C, and 37 degrees C, indicating that the bLH receptor was rotationally immobile on the time scale of our experiments. On these same membrane preparations, the epidermal growth factor (EGF) receptor occupied by ErITC-murine EGF exhibited temperature-dependent rotational correlation times of 80 +/- 5 microseconds, 111 +/- 7 microseconds, 254 +/- 4 microseconds, and > 1000 microseconds at 4 degrees C, 15 degrees C, 25 degrees C, and 37 degrees C, respectively. Slower rotational times for EGF receptor observed at higher temperatures suggested the occurrence of temperature-dependent receptor aggregation. Like the bLH receptor, the oLH receptor on intact cells and on CL plasma membranes was rotationally immobile on the time scale of our experiments when occupied by ErITC-hCG. However, the oLH-occupied receptors on small luteal cells and on luteal cell membranes had comparable rotational correlation times at 37 degrees C. These results suggest that bLH receptors are present in large, rotationally immobile structures, whereas the receptor-containing structure formed on ovine luteal cells depends on whether that receptor is occupied by hCG or oLH. Also, despite the similarities between reproductive function in these species, the LH-occupied receptor appears to be organized differently in the plasma membranes of these hormone-responsive luteal cells.

Flow cytometric analysis of T-independent antigen binding to dinitrophenyl-specific cells

Binding of Ag to membrane Ig (mIg) can lead to either activation or desensitization of the B cell. For thymus-independent (TI) Ags the nature and concentration of the Ag determines what type of signal is delivered to the cell. These Ags are capable of directly activating B lymphocytes and are an important model system for the study of mechanisms involved in B cell responses. In this study, we quantified TI Ag binding and B cell receptor involvement as functions of TI Ag structure, concentration, and epitope density. Various epitope densities of two structurally different TI Ags, DNP-polymerized flagellin (pol) and DNP-dextran (dex), were labeled with tetramethylrhodamine isothiocyanate (TRITC) and reacted with DNP-specific murine splenic B lymphocytes and with cells of a cloned DNP-specific cell line. The amount of Ag bound to the cell surface at various doses was measured directly by flow cytometry. For each Ag and dose, FITC-labeled DNP-L-papain was used to quantitate receptor sites not occupied by Ag. Approximately 5% receptor occupancy was observed for immunogenic doses of Ag. Higher Ag concentrations that can induce tolerance caused a substantial increase in the fraction of occupied receptors. This suggests that tolerogenic responses result from an overly restrictive cross-linking of surface receptors. By comparing these data to previously published data on biologic activity of the Ags, we are able to more clearly define those conditions of Ag binding that lead to B cell activation.

Flow cytometric analysis of T-independent antigen binding to dinitrophenyl-specific cells

Binding of Ag to membrane Ig (mIg) can lead to either activation or desensitization of the B cell. For thymus-independent (TI) Ags the nature and concentration of the Ag determines what type of signal is delivered to the cell. These Ags are capable of directly activating B lymphocytes and are an important model system for the study of mechanisms involved in B cell responses. In this study, we quantified TI Ag binding and B cell receptor involvement as functions of TI Ag structure, concentration, and epitope density. Various epitope densities of two structurally different TI Ags, DNP-polymerized flagellin (pol) and DNP-dextran (dex), were labeled with tetramethylrhodamine isothiocyanate (TRITC) and reacted with DNP-specific murine splenic B lymphocytes and with cells of a cloned DNP-specific cell line. The amount of Ag bound to the cell surface at various doses was measured directly by flow cytometry. For each Ag and dose, FITC-labeled DNP-L-papain was used to quantitate receptor sites not occupied by Ag. Approximately 5% receptor occupancy was observed for immunogenic doses of Ag. Higher Ag concentrations that can induce tolerance caused a substantial increase in the fraction of occupied receptors. This suggests that tolerogenic responses result from an overly restrictive cross-linking of surface receptors. By comparing these data to previously published data on biologic activity of the Ags, we are able to more clearly define those conditions of Ag binding that lead to B cell activation.

Rotational dynamics of luteinizing hormone receptors and MHC class I antigens on murine Leydig cells

We have examined the molecular motions of luteinizing hormone (LH) receptor and the Major Histocompatibility Complex Class I antigen on murine Leydig cells. Using time-resolved phosphorescence anisotropy methods, erythrosin (ErITC)-derivatized ovine luteinizing hormone (oLH) bound to the LH receptor appears rotationally mobile with rotational correlation times of 19.6 +/- 1.3 microseconds, 13.3 +/- 2.4 microseconds, 9.5 +/- 0.7 microseconds and 4.7 +/- 0.5 microseconds at 4 degrees C, 15 degrees C, 25 degrees C and 37 degrees C, respectively. Rotational correlation times for human chorionic gonadotropin (hCG)-occupied LH receptors were similar to those of the ErITC-oLH occupied receptor at each temperature. In addition, both oLH- and hCG-occupied LH receptors were laterally mobile in fluorescence photobleaching recovery experiments with diffusion coefficients at 29 degrees C of (5.8 +/- 0.9) x 10(-10) cm2 s-1 and (2.9 +/- 0.4) x 10(-10) cm2 s-1, respectively. We also measured the rotational correlation time of Class I antigen on murine Leydig cells using ErITC-derivatized 34-12-2S, an anti-Class I monoclonal antibody. Because there was no decay of the anisotropy function at 4 degrees C, 15 degrees C, 25 degrees C or 37 degrees C in the absence of oLH or following preincubation of Leydig cells with 1 nM oLH, it appears that Class I is rotationally immobile on the 1 ms timescale of our experiments. This result is consistent with the presence of Class I antigen in large molecular weight structures and may be the result of Class I self-aggregation. Further, treatment of cells with anti-Class I antibody had no effect on either basal or oLH-stimulated testosterone secretion. Thus, it appears that this anti-Class I antibody is not LH-mimetic on murine Leydig cells.

Truncation of the A alpha chain of MHC class II molecules results in inefficient antigen presentation to antigen-specific T cells

Antigen presenting cells (APC) expressing MHC class II molecules composed of chains with part or all of the cytoplasmic domains deleted are inefficient at presenting hen egg lysozyme peptides to antigen specific T cell hybrids compared with APC that express wild-type MHC class II molecules. This effect is most apparent for mutants in which the alpha chain has been truncated. The inefficiency in antigen presentation can be amplified by pulsing the APC for 4 h with peptide rather than having peptide present throughout the presentation assay. Fixation of antigen-pulsed APC improves the capacity of APC with truncated class II molecules to stimulate T cell hybrids. Fixation of APC prior to exposure to antigen also leads to significant improvement in antigen presentation by the truncated class II molecules. Because the inefficiency of a given hybrid for antigen presentation does not correlate with its ability to transduce a signal as measured by protein kinase C translocation, we suggest that defects in this pathway are not the only cause of impaired antigen presentation. However, because previous studies have demonstrated the need for an intact cytoskeleton for successful antigen presentation, we propose that the carboxy truncated class II molecules are inefficient in antigen presentation because they are unable to generate the signal that ultimately leads to their interaction with the cytoskeleton. These observations underscore the complexity of the events that are required for achieving effective interactions between MHC class II molecules and TCR, and suggest, with regard to efficient antigen presentation, that the physical state of the class II molecules is at least as important as their signal transducing capacity.

Quantitation of fluorescence energy transfer between cell surface proteins via fluorescence donor photobleaching kinetics

We describe practical aspects of photobleaching fluorescence energy transfer measurements on individual living cells. The method introduced by T. M. Jovin and co-workers (see, most recently, Kubitscheck et al. 1993. Biophys. J. 64:110) is based on the reduced rate of irreversible photobleaching of donor fluorophores when acceptor fluorophores are present. Measuring differences in donor photobleaching rates on cells labeled with donor only (fluorescein isothiocyanate-conjugated proteins) and with both donor and acceptor (tetramethylrhodamine-conjugated proteins) allows calculation of the fluorescence energy transfer efficiency. We assess possible methods of data analysis in light of the underlying processes of photobleaching and energy transfer and suggest optimum strategies for this purpose. Single murine B lymphocytes binding various ratios of donor and acceptor conjugates of tetravalent concanavalin A (Con A) and divalent succinyl Con A were examined for interlectin energy transfer by these methods. For Con A, a maximum transfer efficiency of 0.49 +/- 0.02 was observed. Under similar conditions flow cytometric measurements of donor quenching yielded a value of 0.54 +/- 0.03. For succinyl Con A, the maximum transfer efficiency was 0.36. To provide concrete examples of quantities arising in such energy transfer determinations, we present examples of individual cell data and kinetic analyses, population rate constant distributions, and error estimates for the various quantities involved.

Class I major histocompatibility complex antigens are not associated with the LH/CG receptor on ovine luteal cells

We have examined the rotational dynamics of the luteinizing hormone (LH) receptor on day 10 intact ovine small luteal cells and isolated plasma membranes using polarized fluorescence depletion (PFD). This technique measures rotational correlation times which are proportional to the in-membrane volume of a protein and are useful for examining changes in protein size due to receptor aggregation or protein-protein interactions. Eosin isothiocyanate (EITC)-derivatized ovine LH (EITC-oLH) bound to the LH receptor on luteal cell plasma membranes had a rotational correlation time of 20 +/- 6 microseconds, while that for EITC-human chorionic gonadotropin (EITC-hCG)-occupied LH receptors was 46 +/- 13 microseconds. Slower rotational times for EITC-oLH and EITC-hCG, 63 +/- 19 and 87 +/- 20 microseconds, respectively, were obtained on intact ovine luteal cells. These results indicate that the LH receptor exists as a larger molecular mass complex when binding hCG than oLH, a difference which could be attributable to hCG-induced LH-receptor interaction with additional membrane protein(s). One candidate protein for such an interaction is the Major Histocompatibility Complex (MHC) Class-I antigen. However, the rotational correlation time of EITC-anti-MHC Class-I antibody (SBU I) Fab fragments was 247 +/- 34 microseconds, indicating that MHC Class I is located in complexes larger than those identified by EITC-OLH or EITC-hCG. Preincubation of plasma membranes with 1 nM unlabeled oLH or hCG had no significant effect on this rotational correlation time. Further, treatment of cells with SBU I had no affect on either basal or oLH-stimulated progesterone secretion. Thus it appears that the ovine luteal LH-receptor is not associated with MHC Class I and that antibody-induced aggregation of MHC Class I does not cause an LH-mimetic response.

Fluorescence depletion measurements in various experimental geometries provide true emission and absorption anisotropies for the study of protein rotation

Use of fluorescence depletion methods for measuring slow protein rotational diffusion has been limited by failure to obtain, from depletion data, well-defined anisotropy functions dependent on the distribution of either fluorophore emission or absorption transition dipoles, but not both. Such anisotropies would be directly comparable to those obtained from phosphorescence emission or triplet absorption measurements. We now describe such procedures applicable to cuvet and microscope experimental geometries, together with supporting experimental results. In cuvet measurements, the pump and probe beams are colinear and fluorescence is collected at 90 degrees to this axis. The data analysis procedure for this geometry has been suggested by Wegener (Biophys. J., 46 (1984) 795) and permits calculation of the absorption and emission anisotropies and the interdipole angle. In microscope experiments, fluorescence emission is collected along the pump/probe beam axis. For microscope measurements, a new experimental procedure permits evaluation of absorption and emission anisotropies when the interdipole angle is independently known. In either case multiple depletion measurements are required, each with different relative orientations of the probe beam polarization, pump beam polarization and emission polarizer axis. We have used these methods to calculate the time-dependent anisotropies for eosin-derivatized BSA rotation in glycerol solutions in both experimental geometries. These data correspond well with those obtained from time-resolved phosphorescence anisotropy measurements.

Membrane changes in lipopolysaccharide-stimulated murine B lymphocytes associated with cell activation

The lateral diffusion of the fluorescent lipid analog 3,3'-dioctadecylindocarbocyanine iodide (DiI) was measured in the membranes of murine B lymphocytes treated with the B cell mitogen lipopolysaccharide (LPS). The mobility of DiI, as measured by fluorescence photobleaching recovery (FPR) techniques, was temperature-dependent with a value of 6.1.10(-9) cm2 s-1 at 37 degrees C. Untreated cells exhibited this diffusion coefficient over 72 h in culture. In contrast, DiI mobility decreased to 2.0.10(-9) cm2 s-1 at 37 degrees C in membranes of LPS-stimulated lymphocytes 24 h following LPS exposure. Interestingly, this decreased lipid lateral diffusion was not accompanied by any change in surface immunoglobulin lateral diffusion which remained essentially unchanged at 3.6-4.3.10(-11) cm2 s-1 over 72 h. To determine whether LPS effects on lipid lateral diffusion were due to insertion of LPS into the cell plasma membrane, we examined TRITC-LPS diffusion in B lymphocytes from LPS-responsive Balb/c and C3Heb/FeJ mice and from hypo-responsive C3H/HeJ mice. DiI and TRITC-LPS mobility decreased more than 50% in LPS-stimulated Balb/c and C3Heb/FeJ cells by 72 h. On C3H/HeJ lymphocytes, there was no change in DiI or TRITC-LPS lateral diffusion throughout the incubation period. These data indicate that B lymphocyte membrane composition is altered in LPS-activated lymphoblasts and that the decreased lateral diffusion of lipid probes does not result from membrane perturbation by LPS insertion into the lipid bilayer. Further, similarities between TRITC-LPS and DiI lateral diffusion suggest that most LPS molecules interact non-specifically with B cell membranes, presumably by acyl chain insertion of the lipid A moiety.

Molecular dynamics of luteinizing hormone receptors on rat luteal cells

To better understand the in situ organization of the luteinizing hormone receptor on rat luteal cells, we have examined the molecular motions of this receptor following binding of ovine luteinizing hormone (oLH) or human chorionic gonadotropin (hCG). Fluorescence photobleaching recovery (FPR) measurements of LH receptor lateral diffusion were performed using tetramethylrhodamine isothiocyanate (TRITC)-derivatized oLH or hCG as a probe. These experiments indicate that TRITC-oLH occupied LH receptors on luteal cells obtained from superovulated female rats have a lateral diffusion coefficient D of (1.7 +/- 0.6).10(-10) cm2s-1 at 27 degrees C with fluorescence recovery after photobleaching of 46 +/- 5%. In similar experiments, binding of TRITC-hCG caused a significant decrease in LH receptor lateral diffusion; fluorescence recovery after photobleaching was less than 20%. To determine whether hCG-occupied receptors might exist in large aggregates, we measured the rotational correlation times (RCT) of hCG and oLH bound to the LH receptor on intact cells using single cell polarized fluorescence depletion (PFD). At 4 degrees C, LH receptors occupied by eosin isothiocyanate (EITC)-hCG exhibited a slower RCT (64 microseconds) than did receptors occupied by EITC-oLH (43 microseconds). At this temperature both TRITC-oLH and TRITC-hCG occupied LH receptors were laterally immobile. These FPR and PFD results suggest that the molecular motions of the luteal cell LH receptor are significantly modulated by the subtle structural differences in various bound gonadotropins.

Self-association of class I major histocompatibility complex molecules in liposome and cell surface membranes

Fluorescent derivatives of a human MHC class I glycoprotein, HLA-A2, were reconstituted into dimyristoylphosphatidylcholine (DMPC) liposomes. Measurements of lateral diffusion of fluorescein-(Fl-) labeled HLA-A2 by fluorescence photobleaching recovery (FPR), of rotational diffusion of erythrosin-(Er-) labeled HLA-A2 by time-resolved phosphorescence anisotropy (TPA), and of molecular proximity by flow cytometric fluorescence resonance energy transfer (FCET) showed that these class I MHC molecules self-associate in liposome membranes, forming small aggregates even at low surface concentrations. The lateral diffusion coefficient (Dlat) of Fl-HLA-A2 decreases with increasing surface protein concentration over a range of lipid:protein molar ratios (L/P) between 8000:1 and 2000:1. The reduction in Dlat of HLA molecules in DMPC liposomes is found to be sensitive to time and temperature. The rotational correlation time for Er-HLA-A2 in DMPC liposomes at 30 degrees C is 87 +/- 0.8 microseconds, at least 10 times larger than that expected for an HLA monomer. There is also significant quenching of donor (Fl-HLA) fluorescence at 37 degrees C in the presence of acceptor-labeled (sulforhodamine-labeled HLA) protein indicating proximity between HLA molecules even at L/P = 4000:1. FPR and FCET measurements with another membrane glycoprotein, glycophorin, give no evidence for its self-association. HLA aggregation measured by FPR, FCET, and TPA was blocked by beta 2-microglobulin, b2m, added to the liposomes. The aggregation of HLA-A2 molecules is not an artifact of their reconstitution into liposomes. HLA aggregates, defined by FCET, were readily detected on the surface of human lymphoblastoid (JY) cells.(ABSTRACT TRUNCATED AT 250 WORDS)

Protein and lipid lateral diffusion in normal and Rous sarcoma virus transformed chick embryo fibroblasts

We measured the lateral diffusion of the fluorescent lipid analogue dioctadecylindocarbocyanine iodide (DiI) and of membrane glycoproteins labeled with tetramethylrhodamine (TRITC) succinyl concanavalin A (SConA) via fluorescence photobleaching recovery (FPR) at selected times during a temperature downshift experiment on transformation-defective temperature-sensitive (td-ts) Rous sarcoma virus (RSV) NY68-transformed chicken embryo fibroblasts (CEF) and on identically treated CEF and RSV-transformed CEF. There were no significant differences in the lateral diffusion in DiI at any of the times measured. The lateral diffusion of TRITC-SConA on the RSV-transformed CEF, (1.32 +/- 0.12).10(-10) cm2 s-1, was approximately two times faster than that observed in normal CEF, (0.61 +/- 0.06).10(-10) cm2 s-1. In the cells undergoing RSV NY68-mediated transformation, TRITC-SConA diffusion increased over a 24-h period from a value comparable to that observed in normal CEF, (0.72 +/- 0.13).10(-10) cm2 s-1 to a value comparable to the RSV-CEF transformed cells, (1.74 +/- 0.20).10(-10) cm2 s-1. All diffusion measurements reported were made at the permissive temperature for RSV-NY68 (35 degrees C) unless stated otherwise. The changes in the lateral diffusion of TRITC-SConA occurred between the fifth and twelfth hour of the downshift course and could be associated with cytoskeletal disruption and/or fibronectin degradation, both known to occur at this time in RSV-transformed cells. To assess the contribution of extracellular matrix (ECM) degradation, SConA mobility was measured in normal and RSV-transformed cells treated with trypsin. This treatment increased SConA mobility approximately 4-fold in the normal cells relative to untreated controls and only 2-fold in the RSV-CEF transformed cells. No significant difference in SConA mobility between trypsinized spherical normal and transformed cells was apparent.

Rotational dynamics of type I Fc epsilon receptors on individually-selected rat mast cells studied by polarized fluorescence depletion

We report the first application of polarized fluorescence depletion (PFD), a technique which combines the sensitivity of fluorescence detection with the long lifetimes of triplet probes, to the measurement of membrane protein rotational diffusion on individually selected, intact mammalian cells. We have examined the rotation of type I Fc epsilon receptors (Fc epsilon RI) on rat mucosal mast cells of the RBL-2H3 line in their resting monomeric and differently oligomerized states using as probes IgE and three monoclonal antibodies (mAbs; H10, J17, and F4) specific for the Fc epsilon RI. PFD experiments using eosin (EITC)-IgE show that individual Fc epsilon RI on cells have a rotational correlation time (RCT) at 4 degrees C of 79 +/- 4 microseconds. Similarly, Fc epsilon RI-bound EITC-Fab fragments of the J17 Fc epsilon RI-specific mAb exhibit an RCT of 76 +/- 6 microseconds. These values agree with previous measurements of Fc epsilon RI-bound IgE rotation by time-resolved phosphorescence anisotropy methods. Receptor-bound EITC-conjugated divalent J17 antibody exhibits an increased RCT of 140 +/- 6 microseconds. This is consistent with the ability of this mAb to form substantial amounts of Fc epsilon RI dimers on these cell surfaces. The ratio of limiting to initial anisotropy in these experiments remains constant at about 0.5 from 5 degrees C through 25 degrees C for IgE, Fab, and intact mAb receptor ligands. Extensive cross-linking by second antibody of cell-bound IgE, of intact Fc epsilon RI-specific mAbs or of their Fab fragments, however, produced large fixed anisotropies demonstrating, under these conditions, receptor immobilization in large aggregates. PFD using the mAbs H10 and F4 as receptor probes yielded values for triplet lifetimes, RCT values, and anisotropy parameters essentially indistinguishable from those obtained with the mAb J17 clone. Possible explanations for these observations are discussed.

Triamcinolone acetonide inhibits lymphocyte differentiation in B cells decorated with artificial antigen receptors

We examined the effects of triamcinolone acetonide (TA) on T cell independent antigen-induced differentiation of human B cells. Purified human B cells artificially decorated with palmitate-conjugated monoclonal IgA antibody specific for 2,4-dinitrophenyl differentiated polyclonally when challenged with optimum concentrations of dinitrophenyl-derivatized polymerized flagellin. This B cell response was reduced by the synthetic corticosteroid TA at a concentration of 10(-6) M. This suggests that TA can inhibit in vitro B lymphocyte differentiation independent of T cells.

Corticosteroid effects on lipid lateral diffusion in CEM-C1 and CEM-C7 acute lymphoblastic leukemia cells

We have examined glucocorticoid effects on CEM-C7 and CEM-C1 subclones of a leukemic human T-cell line using fluorescence photobleaching recovery techniques. Incubation with 10(-5) M triamcinolone acetonide (TA) increased lipid lateral diffusion on steroid-sensitive CEM-C7 cells but had no effect on steroid-resistant CEM-C1 cells. CEM-C7 cells incubated in serum-free medium responded only to TA but, when fetal calf serum was added to the incubation medium, would also respond to 10(-5) M dexamethasone and hydrocortisone. Thus, glucocorticoids can cause increased lipid lateral diffusion in CEM-C7 cells, while having no effect on steroid-resistant CEM-C1 cells.

Antigen activation of human B lymphocytes bearing artificial antigen receptors

When highly purified human and murine B cells are challenged in vitro with certain so called "T cell-independent" activators such as the polyclonal B cell activator lipopolysaccharide (LPS) or the clonally specific B cell activator dinitrophenyl-conjugated polymerized flagellin (DNP-POL), mouse, but not human, cells differentiate into immunoglobulin-secreting cells. However, results from this study show that DNP-POL can cause human B cell differentiation in a T cell-independent manner when the antigen is concentrated onto the cells via artificially incorporated palmitate-modified anti-DNP mouse IgA molecules. This response is comparable in magnitude to that induced by a T cell-dependent polyclonal B cell activator, pokeweed mitogen, in unfractionated mononuclear cell cultures, suggesting that DNP-POL induced polyclonal B cell differentiation. DNP-POL binding to the artificial receptor molecules on B cells did not cause cellular proliferation, even in unfractionated mononuclear cell populations. These results are similar to those obtained in previous studies using mouse B cells in which the artificial receptor was unable to act as a transmembrane signaling element. From these studies, we conclude that B cells express clonally unrestricted, presumably low-avidity, endogenous receptor for POL, and that signaling through this receptor activates B cell differentiation but not cell proliferation.