Anomalous diffusion of major histocompatibility complex class I molecules on HeLa cells determined by single particle tracking.

Single-particle tracking (SPT) was used to determine the mobility characteristics of MHC (major histocompatibility complex) class I molecules at the surface of HeLa cells at 22 degrees C and on different time scales. MHC class I was labeled using the Fab fragment of a monoclonal antibody (W6/32), covalently bound to either R-phycoerythrin or fluorescent microspheres, and the particles were tracked using high-sensitivity fluorescence imaging. Analysis of the data for a fixed time interval suggests a reasonable fit to a random diffusion model. The best fit values of the diffusion coefficient D decreased markedly, however, with increasing time interval, demonstrating the existence of anomalous diffusion. Further analysis of the data shows that the diffusion is anomalous over the complete time range investigated, 4-300 s. Fitting the results obtained with the R-phycoerythrin probe to D = D0talpha-1, where Do is a constant and t is the time, gave D0 = (6.7 +/- 4.5) x 10(-11) cm2 s-1 and alpha = 0.49 +/- 0.16. Experiments with fluorescent microspheres were less reproducible and gave slower anomalous diffusion. The R-phycoerythrin probe is considered more reliable for fluorescent SPT because it is small (11 x 8 nm) and monovalent. The type of motion exhibited by the class I molecules will greatly affect their ability to migrate in the plane of the membrane. Anomalous diffusion, in particular, greatly reduces the distance a class I molecule can travel on the time scale of minutes. The present data are discussed in relation to the possible role of diffusion and clustering in T-cell activation.     

Lowering the barriers to random walks on the cell surface

We used fluorescence recovery after photobleaching (FRAP) and single particle tracking (SPT) techniques to compare diffusion of class I major histocompatibility complex molecules (MHC) on normal and alpha-spectrin-deficient murine erythroleukemia (MEL) cells. Because the cytoskeleton mesh acts as a barrier to lateral mobility of membrane proteins, we expected that diffusion of membrane proteins in alpha-spectrin-deficient MEL cells would differ greatly from that in normal MEL cells. In the event, diffusion coefficients derived from either FRAP or SPT analysis were similar for alpha-spectrin-deficient and normal MEL cells, differing by a factor of approximately 2, on three different timescales: tens of seconds, 1-10 s, and 100 ms. SPT analysis showed that the diffusion of most class I MHC molecules was confined on both cell types. On the normal MEL cells, the mean diagonal length of the confined area was 330 nm with a mean residency time of 40s. On the alpha-spectrin-deficient MEL cells, the mean diagonal length was 650 nm with a mean residency time of 45s. Thus there are fewer barriers to lateral diffusion on cytoskeleton mutant MEL cells than on normal MEL cells, but this difference does not strongly affect lateral diffusion on the scales measured here.     

Lateral mobility of class I histocompatibility antigens in B lymphoblastoid cell membranes: modulation by cross-linking and effect of cell density

We have studied the lateral mobility of class 1 major histocompatibility complex (MHC) proteins in the membranes of human Epstein-Barr virus-transformed B cells using fluorescence photobleaching recovery. Class I MHC antigens were labeled with either W6/32 monoclonal antibody or its Fab fragment directly conjugated to fluorescein isothiocyanate. The diffusion coefficient of class I antigens labeled with Fab fragments of W6/32 was identical to that of a lipid analogue, fluorescein phosphatidylethanolamine, and was 10-fold greater than that of antigens labeled with intact W6/32. Furthermore, antigens labeled with Fab fragments but not with intact W6/32 had fractional mobilities identical to that of the lipid probe. The lateral mobility of class I antigens was dependent on the time of incubation with fluorescent antibody and on the presence of antibody microaggregates. Finally, class I MHC proteins labeled with intact W6/32 but not with Fab fragments were immobilized in the membranes of most cells grown in suspension at high cell density. These results suggest that, in the unperturbed state, class I MHC antigens diffuse as rapidly as membrane lipid, i.e., without cytoskeletal constraint. Cross-linking with bivalent ligand and growth to high cell density may trigger membrane events leading to slowing and immobilization of these proteins.     

Membrane-bound and water-soluble nonclassical class I MHC antigens on rat placenta

We have used mouse monoclonal antibodies to different determinants on rat class I major histocompatibility complex (MHC) antgiens in order to identify water-soluble and membrane-bound nonclassical (i.e., non-RT1.A) class I MHC antigens on the spongiotrophoblast and labyrinthine trophoblast of rat placenta. Initial immunohistological studies with monoclonal antibodies reacting with determinant restricted to classical (RT1.A) rat class I antigens confirmed the presence of these antigens on spongiothrophoblast, but not on labyrinthine trphoblast. Staining with another monoclonal antibody, which reacts with both classical and at least some nonclassical rat class I antigens, gave strong staining of both the labyrinthine and spongiotrophoblast. To distinguish membrane-bound from water-soluble class I molecules, quantitative adsorption analyses were carried out using both placental cell membranes and ultracentrifuged aqueous extracts of placenta. The aqueous placental extract had no absorptive capacity for the RT1.A-specific antibodies, but it had very strong absorptive capacity for the more broadly reactive antibody. This strongly suggests the presence of large quantities of a soluble nonclassical class I MHC antigen in rat placenta. The placental cell membranes had four to fivefold greater absorptive capacity for the broadly reactive antibody when compared to the antibodies to classical class I antigens, a result that was consistent with the presence of membrane-bound non-classical class I antigens on rat placenta. The membrane-bound nonclassical class I antigen was purified from detergent extracts of DA rat placental membranes using monoclonal antibody affinity and lentil lectin affinity chromatography. The putative nonclassical class I antigen had a heavy chain of M _r 43 000, which is 2000 smaller than the amino acid sequence analysis demonstrated that the nonclassical placental antigen differed at three amino acid residues from the classical RT1.A class I molecule and also from the Q10-like class I molecule of the DA strain. It differed also from the pAR 1.5 cDNA sequence, the only full-length rat class I DNA sequence available so far. Address correspondence and offprint requests to: J. Fabre.     

Invariant chain and the MHC class II cytoplasmic domains regulate localization of MHC class II molecules to lipid rafts in tumor cell-based vaccines

Cell-based tumor vaccines, consisting of MHC class I+ tumor cells engineered to express MHC class II molecules, stimulate tumor-specific CD4+ T cells to mediate rejection of established, poorly immunogenic tumors. Previous experiments have demonstrated that these vaccines induce immunity by functioning as APCs for endogenously synthesized, tumor-encoded Ags. However, coexpression of the MHC class II accessory molecule invariant chain (Ii), or deletion of the MHC class II cytoplasmic domain abrogates vaccine immunogenicity. Recent reports have highlighted the role of lipid microdomains in Ag presentation. To determine whether Ii expression and/or truncation of MHC class II molecules impact vaccine efficacy by altering MHC class II localization to lipid microdomains, we examined the lipid raft affinity of MHC class II molecules in mouse M12.C3 B cell lymphomas and SaI/A(k) sarcoma vaccine cells. Functional MHC class II heterodimers were detected in lipid rafts of both cell types. Interestingly, expression of Ii in M12.C3 cells or SaI/A(k) cells blocked the MHC class II interactions with cell surface lipid rafts. In both cell types, truncation of either the alpha- or beta-chain decreased the affinity of class II molecules for lipid rafts. Simultaneous deletion of both cytoplasmic domains further reduced localization of class II molecules to lipid rafts. Collectively, these data suggest that coexpression of Ii or deletion of the cytoplasmic domains of MHC class II molecules may reduce vaccine efficacy by blocking the constitutive association of MHC class II molecules with plasma membrane lipid rafts.     

Mobility in the mitochondrial electron transport chain

The role of lateral diffusion in mitochondrial electron transport has been investigated by measuring the diffusion coefficients for lipid, cytochrome c, and cytochrome oxidase in membranes of giant mitoplasts from cuprizone-fed mice using the technique of fluorescence redistribution after photobleaching (FRAP). The diffusion coefficient of the phospholipid analogue N-(7-nitro-2,1,3-benzoxadiazol-4-yl)phosphatidylethanolamine is dependent on the technique used to remove the outer mitochondrial membrane. A sonication technique yields mitoplasts with monophasic recovery of the lipid probe (D = 6 X 10(-9) cm2/s), while digitonin-treated mitochondria show biphasic recoveries (D1 = 5 X 10(-9) cm2/s; D2 = 1 X 10(-9) cm2/s). Digitonin appears to incorporate into mitoplasts, giving rise to decreased lipid mobility concomitant with increased rates of electron transfer from succinate to oxygen, in a manner reminiscent of the effects of cholesterol incorporation [Schneider, H., Lemasters, J. J., Hochli, M., & Hackenbrock, C. R. (1980) J. Biol. Chem. 255, 3748-3756]. FRAP measurements on tetramethylrhodamine cytochrome c modified at lysine-39 and on a mixture of active morpholinorhodamine derivatives of cytochrome c gave diffusion coefficients of (3.5-7) X 10(-10) cm2/s depending on the assay medium. With morpholinorhodamine-labeled antibodies purified on a cytochrome oxidase affinity column, the diffusion coefficient for cytochrome oxidase was determined to be 1.5 X 10(-10) cm2/s. The results are discussed in terms of a dynamic aggregate model in which an equilibrium exists between freely diffusing and associated electron-transfer components.     

Intracellular macromolecular mobility measured by fluorescence recovery after photobleaching with confocal laser scanning microscopes

Fluorescence recovery after photobleaching (FRAP) is a widely used tool for estimating mobility parameters of fluorescently tagged molecules in cells. Despite the widespread use of confocal laser scanning microscopes (CLSMs) to perform photobleaching experiments, quantitative data analysis has been limited by lack of appropriate practical models. Here, we present a new approximate FRAP model for use on any standard CLSM. The main novelty of the method is that it takes into account diffusion of highly mobile molecules during the bleach phase. In fact, we show that by the time the first postbleach image is acquired in a CLSM a significant fluorescence recovery of fast-moving molecules has already taken place. The model was tested by generating simulated FRAP recovery curves for a wide range of diffusion coefficients and immobile fractions. The method was further validated by an experimental determination of the diffusion coefficient of fluorescent dextrans and green fluorescent protein. The new FRAP method was used to compare the mobility rates of fluorescent dextrans of 20, 40, 70, and 500 kDa in aqueous solution and in the nucleus of living HeLa cells. Diffusion coefficients were lower in the nucleoplasm, particularly for higher molecular weight dextrans. This is most likely caused by a sterical hindrance effect imposed by nuclear components. Decreasing the temperature from 37 to 22 degrees C reduces the dextran diffusion rates by approximately 30% in aqueous solution but has little effect on mobility in the nucleoplasm. This suggests that spatial constraints to diffusion of dextrans inside the nucleus are insensitive to temperature.     

Docosahexaenoic acid-containing phosphatidylcholine affects the binding of monoclonal antibodies to purified Kb reconstituted into liposomes

Class I major histocompatibility complex (MHC I) molecules are transmembrane proteins that bind and present peptides to T-cell antigen receptors. The role of membrane lipids in controlling MHC I structure and function is not understood, although membrane lipid composition influences cell surface expression of MHC I. We reconstituted liposomes with purified MHC I (Kb) and probed the effect of lipid composition on MHC I structure (monoclonal anti-MHC I antibody binding). Four phospholipids were compared; each had a phosphocholine head group, stearic acid in the sn-1 position, and either oleic, alpha-linolenic, arachidonic, or docosahexaenoic acid (DHA) in the sn-2 position. The greatest binding of monoclonal antibody AF6-88.5, which detects a conformationally sensitive epitope in the extracellular region of the MHC I alpha-chain, was achieved with DHA-containing proteoliposomes. Other epitopes (CTKb, 5041.16.1) showed some sensitivity to lipid composition. The addition of beta2-microglobulin, which associates non-covalently with the alpha-chain and prevents alpha-chain aggregation, did not equalize antibody binding to proteoliposomes of different lipid composition, suggesting that free alpha-chain aggregation was not responsible for disparate antibody binding. Thus, DHA-containing membrane lipids may facilitate conformational change in the extracellular domains of the alpha-chain, thereby modulating MHC I function through effects on that protein's structure.     

Inferences from FRAP data are model dependent: A subdiffusive analysis

Fluorescence recovery after photobleaching (FRAP) is a widely used biological experiment to study the kinetics of molecules that react and move randomly. Since the development of FRAP in the 1970s, many reaction-diffusion models have been used to interpret FRAP data. However, intracellular molecules are widely observed to move by anomalous subdiffusion instead of normal diffusion. In this article, we extend a popular reaction-diffusion model of FRAP to the case of subdiffusion modeled by a fractional diffusion equation. By analyzing this reaction-subdiffusion model, we show that FRAP data are consistent with both diffusive and subdiffusive motion in many scenarios. We illustrate this general result by fitting our model to FRAP data from glucocorticoid receptors in a cell nucleus. We further show that the assumed model of molecular motion (normal diffusion or subdiffusion) strongly impacts the biological parameter values inferred from a given experimentally observed FRAP curve. We additionally analyze our model in three simplified parameter regimes and discuss parameter identifiability for varying subdiffusion exponents.     

The effects of Ca2+ on lipid diffusion

The effects of Ca2+ on rotational and translational diffusion of lipids in multilamellar dimyristoylphosphatidylcholine (DMPC)-water systems were investigated by time-resolved phosphorescence anisotropy steady-state fluorescence polarization and fluorescence recovery after photobleaching (FRAP) experiments. Above the phase transition temperature (Tm), addition of Ca2+ caused a steady increase in the segmental motion of the phosphorescent probe, but resulted in slower diffusion of the fluorescent and lateral diffusion probes. The former result is attributed to changes in the structure of the lipid/water interface that affects the chromophore mobility on the phosphorescence time scale but does not reflect lipid motion. Below the phase transition temperature, slower diffusion of all probes were observed with increasing concentrations of Ca2+, with sudden large changes occurring at [Ca2+] approximately 500 mM. This behaviour is attributed to association of Ca2+ with the lipid phosphate groups and the exclusion of water molecules which results in tighter packing of lipids and smaller segmental motion, leading eventually to the immobilization of lipid molecules.     

Adenovirus type 5 fiber knob binds to MHC class I alpha2 domain at the surface of human epithelial and B lymphoblastoid cells.

Adenovirus serotype 5 (Ad5) fiber receptor was investigated using reverse antibody biopanning of a phage-displayed hexapeptide library, and virus-neutralizing monoclonal antibodies (mAbs 1D6.3 and 7A2.7) raised against recombinant Ad5 fiber knob. Both mAbs inhibited attachment of Ad5 to HeLa cells. Mimotopes of 1D6.3 showed homology with the C-terminal segment of the alpha2 domain of the heavy chain of human MHC class I molecules (MHC-I alpha2), and mimotopes of 7A2.7 were consensus to human fibronectin type III (FNIII) modules. In vitro, GST-fused MHC-I alpha2- and FNIII-derived oligopeptides interacted with recombinant fibers in a subgroup-specific manner. In vivo, the MHC-I alpha2 synthetic icosapeptide RAIVGFRVQWLRRYFVNGSR showed a net neutralization effect on Ad5 in HeLa cells, whereas the FNIII icosapeptide RHILWTPANTPAMGYLARVS significantly increased Ad5 binding to HeLa cells. Daudi cells, which lack surface expression of HLA class I molecules, showed a weak capacity for Ad5 binding. In beta2-microglobulin-transfected Daudi cells, Ad5 attachment and permissivity were restored to HeLa cell levels, with 4000 receptors per cell and a binding constant of 1.4x10(10)/M. The results suggested that the conserved region of MHC-I alpha2-domain including Trp167 represents a high affinity receptor for Ad5 fiber knob, whereas ubiquitous FNIII modules would serve as auxiliary receptors.     

Cell-specific constraints to the lateral diffusion of a membrane glycoprotein.

We have previously shown that the lateral diffusion, D, of the class I Major Histocompatibility Complex (MHC) glycoprotein H-2Ld is constrained by its glycosylation, when expressed in mouse L-cells. Removal of one or more of the 3 N-linked oligosaccharides of H-2Ld glycoproteins results in an increase in D. In order to further examine the influence of glycosylation on D, we compared lateral diffusion of H-2Ld expressed in wild-type CHO cells with lateral diffusion of the same molecule expressed in mutant CHO cells with aberrant surface glycosylation. In addition, we compared lateral diffusion of wild-type and unglycosylated H-2Ld antigens in these cells. In contrast to the large effect of glycosylation state on lateral diffusion of H-2Ld in mouse L-cells, there was little effect of glycosylation on lateral diffusion of H-2Ld in any of the CHO cells. This, together with similar results on hamster class I antigens, indicates that the constraints to D of H-2Ld and other class I MHC molecules are different in CHO cells than in L-cells. Measurements of lateral diffusion after treatment of cells with cytochalasin D make it clear that interactions between MHC class I molecules and a cytoskeleton are important in reducing the mobile fraction of diffusing molecules, R, though they cannot be shown to directly affect the diffusion coefficient, D.     

Comparison of phosphorylation and internalization of the antigen receptor/CD3 complex, CD8, and class I MHC-encoded proteins on T cells. Role of intracytoplasmic domains analyzed with hybrid CD8/class I molecules

We analyzed the phosphorylation and the dynamics of TCR/CD3, CD8 and MHC class I molecules during the activation of a CD8+ cytotoxic T lymphocyte clone and of CD8- T helper hybridomas transfected with the gene coding for the native (J. Gabert, C. Langlet, R. Zamoyska, J.R. Parnes, A.M. Schmitt-Verhulst, and B. Malissen. 1987. Reconstitution of MHC class I specificity by transfer of the T cell receptor and Lyt-2 genes. Cell 50:545) or truncated CD8 alpha molecule. The CD3 components gamma and epsilon and the CD8 alpha subunit were phosphorylated after activation of the CTL clone with the protein kinase C activator PMA. Class I MHC molecules were phosphorylated irrespective of PMA activation. Constitutive phosphorylation of the MHC class I products was found to be intrinsic to the transmembrane/cytoplasmic portion of the molecules because it was transferred to the CD8 alpha hybrid molecules composed of extracellular CD8 and MHC class I transmembrane and intracytoplasmic domains (CD8-e/MHC-t-i). Measurements of the dynamics of these cell surface molecules by using radiolabeled mAb revealed distinct behaviors: TCR/CD3 complex ligand internalization was increased (around 50% after 40 to 60 min) after PMA activation, whereas the ligand of class I MHC molecules was internalized at constant rate irrespective of PMA activation. Ligand bound to native CD8 molecules was poorly internalized, irrespective of the activation of the T cells with PMA. The same ligand bound to the CD8-e/MHC-t-i hybrid molecule was internalized at the same rate as a class I MHC molecule ligand, indicating that the behavior of the hybrid molecule was characteristic of the transmembrane/cytoplasmic portion of MHC class I molecules.     

Constitutive internalization of murine MHC class I molecules

The total number of cell surface glycoprotein molecules at the plasma membrane results from a balance between their constitutive internalization and their egress to the cell surface from intracellular pools and/or biosynthetic pathway. Constitutive internalization is net result of constitutive endocytosis and endocytic recycling. In this study we have compared spontaneous internalization of murine major histocompatibility complex (MHC) class I molecules (K(d), D(d), full L(d), and empty L(d)) after depletion of their egress to the cell surface (Cycloheximide [CHX], brefeldin A [BFA]) and internalization after external binding of monoclonal antibody (mAb). MHC class I alleles differ regarding their cell surface stability, kinetics, and in the way of internalization and degradation. K(d) and D(d) molecules are more stable at the cell surface than L(d) molecules and, thus, constitutively internalized more slowly. Although the binding of mAbs to cell surface MHC class I molecules results in faster internalization than depletion of their egress, it is still slow and, thereby, can serve as a model for tracking of MHC class I endocytosis. Internalization of fully conformed MHC class I molecules (K(d), D(d), and L(d)) was neither inhibited by chlorpromazine (CP) (inhibitor of clathrin endocytosis), nor with filipin (inhibitor of lipid raft dependent endocytosis), indicating that fully conformed MHC class I molecules are internalized via the bulk pathway. In contrast, internalization of empty L(d) molecules was inhibited by filipin, indicating that non-conformed MHC class I molecules require intact cholesterol-rich membrane microdomains for their constitutive internalization. Thus, conformed and non-conformed MHC class I molecules use different endocytic pathways for constitutive internalization.     

The effect of ligand affinity on integrins’ lateral diffusion in cultured cells

The role of ligand affinity in altering αPS2CβPS integrins’ lateral mobility was studied using single particle tracking (SPT) with ligand-functionalized quantum dots (QDs) and fluorescence recovery after photobleaching (FRAP) with fluorescent protein tagged integrins. Integrins are ubiquitous transmembrane proteins that are vital for numerous cellular functions, including bidirectional signaling and cell anchorage. Wild-type and high ligand affinity mutant (αPS2CβPS-V409D) integrins were studied in S2 cells. As measured by SPT, the integrin mobile fraction decreased by 22 % and had a 4× slower diffusion coefficient for αPS2CβPS-V409D compared to wild-type integrins. These differences are partially the result of αPS2CβPS-V409D integrins’ increased clustering. For the wild-type integrins, the average of all diffusion coefficients measured by SPT was statistically similar to the ensemble FRAP results. A 75 % slower average diffusion coefficient was measured by SPT compared to FRAP for αPS2CβPS-V409D integrins, and this may be the result of SPT measuring only ligand-bound integrins, in contrast all ligand-bound and ligand-unbound integrins are averaged in FRAP measurements. Specific binding of the ligand-functionalized QDs was 99 % for integrin expressing cells. The results prove that the ligand binding affinity affects the lateral dynamics of a subset of integrins based on the complementary SPT and FRAP data.     

Monoclonal antibodies against an HLA-B27-derived peptide react with an epitope present on bacterial proteins.

The role of molecular mimicry in the spondyloarthropathies was investigated with respect to the epitopes involved. mAb were produced against a synthetic peptide whose sequence was derived from a polymorphic region of the HLA-B27 molecule (amino acids 63-83). Two antibodies (J7F2 and H2B6) were selected for study on the basis of their ability to react with bacterial envelope proteins (ELISA) and B27-positive cells (immunofluorescence). J7F2 reacted preferentially with B27-positive cells and neither antibody reacted with MHC class I negative cells. Based on SDS-PAGE blot analysis of bacterial envelope proteins, the pattern of reactivity for both antibodies (against 36- and 19-kDa proteins) was the same as that for a third monoclonal produced against bacterial envelope and reactive with B27-positive cells. This apparent epitope similarity was investigated by using synthetic peptides to inhibit binding of the monoclonals. The B27 synthetic peptide and a smaller peptide derived from it were efficient inhibitors of antipeptide and antibacterial antibody binding to bacterial Ag and B27-positive cells. These studies provide insight into the molecular basis of cross-reactivity between bacterial proteins and MHC class I molecules.     

Lateral diffusion anisotropy and membrane lipid/skeleton interaction in outer hair cells

The organization of the plasma membrane of cells in lipid domains affects the way the membrane interacts with the underlying protein skeleton, which in turn affects the lateral mobility of lipid and protein molecules in the membrane. Membrane fluidity properties can be monitored by various approaches, the most versatile of which is fluorescence recovery after photobleaching (FRAP). We extended previous FRAP experiments on isolated cochlear outer hair cells (OHCs) by analyzing the two-dimensional pattern of lipid diffusion in the lateral membrane of these cells. We found that membrane lipid mobility in freshly isolated OHCs is orthotropic, diffusion being faster in the axial direction of the cell and slower in the circumferential direction. Increasing the cell's turgor pressure by osmotic challenge reduced the axial diffusion constant, but had only a slight effect on circumferential diffusion. Our results suggest that lipid mobility in the OHC plasma membrane is affected by the presence of the cell's orthotropic membrane skeleton. This effect could reflect interaction with spectrin filaments or with other membrane skeletal proteins. We also performed a number of FRAP measurements in temporal bone preparations preserving the structural integrity of the hearing organ. The diffusion rates measured for OHCs in this preparation were in good agreement with those obtained in isolated OHCs, and comparable to the mobility rates measured on the sensory inner hair cells. These observations support the idea that the plasma membranes of both types of hair cells share similar highly fluid phases in the intact organ. Lipid mobility was significantly slower in the membranes of supporting cells of the organ of Corti, which could reflect differences in lipid phase or stronger hindrance by the cytoskeleton in these membranes.     

Differences between the lateral organization of conventional and inositol phospholipid-anchored membrane proteins. A further definition of micrometer scale membrane domains.

Plasma membranes of many cells appear to be divided into domains, areas whose composition and function differ from the average for an entire membrane. We have previously used fluorescence photo-bleaching and recovery to demonstrate one type of membrane domain, with dimensions of micrometers (Yechiel, E., and M. Edidin. 1987, J. Cell Biol. 105: 755-760). The presence of membrane domains is inferred from the dependence of the apparent mobile fraction of labeled molecules on the size of the membrane area probed. We now find that by this definition classical class I MHC molecules, H-2Db, are concentrated in domains in the membranes of K78-2 hepatoma cells, while the nonclassical class I-related molecules, Qa-2, are free to pass the boundaries of these domains. The two proteins are highly homologous but differ in their mode of anchorage to the membrane lipid bilayer. H-2Db is anchored by a transmembrane peptide, while Qa-2 is anchored by a glycosylphosphatidylinositol (GPI) anchor. A mutant class I protein with its external portion derived from Qa-2 but with transmembrane and cytoplasmic sequences from a classical class I molecule shows a dependence of its mobile fraction on the area of membrane probed, while a mutant whose external portions are a mixture of classical and nonclassical class I sequences, GPI-linked to the bilayer, does not show this dependence and hence by our definition is not restricted to membrane domains.     

Measuring rotational diffusion of MHC class I on live cells by polarized FPR

Clustering of membrane proteins is a dynamic process which can regulate cellular function and signaling. The size of receptor and other membrane protein clusters can in principle be measured in terms of their rotational diffusion. However, in practice, measuring rotation of membrane proteins of live cells has been difficult, largely because of the difficulty of rigidly attaching reporter groups to the molecules of interest. Here we show that polarized photobleaching recovery can detect rotation of membrane proteins genetically tagged with yellow fluorescent protein, YFP. MHC class I molecules were engineered with a rigid, in-sequence, YFP tag followed at the C-terminus by a pair of crosslinkable domains. When crosslinker was added we could detect changes in rotational anisotropy decay consistent with clustering of the MHC molecules. This result points the way to use of engineered fluorescent fusion proteins to measure rotational diffusion in native cell membranes.