Search

It has been suggested that products of the major histocompatibility complex, the MHC, of vertebrates function in many processes of recognition and ligand binding at the cell surface. Here we show that binding of polyclonal and monoclonal antibodies against human MHC antigens, HLA, reduced the binding of epidermal growth factor (EGF) to its membrane receptors on A-431 tumor cells and on normal human fibroblasts. Binding of EGF at 37 degrees C similarly inhibited the binding of Fab fragments and intact Ig anti-HLA to human cells. The inhibitory effect of anti-HLA antibodies was rapid and dependent upon temperature and antibody concentration and valence. Fluorescence microscopy qualitatively confirmed the binding data and showed that MHC antigens and EGF-receptors do not co-cluster in the membrane.

We have used isolated embryonic photoreceptor cells as a model system with which to examine the mechanisms responsible for the development and maintenance of asymmetric Na+,K+-ATPase (ATPase) distribution. Photoreceptor precursors, which appear round and process free at culture onset, develop structural and molecular properties similar to those of photoreceptor cells in vivo. ATPase, recognized by an anti-ATPase antibody, is distributed over the entire surface of round photoreceptor precursors. As the cells develop, ATPase becomes progressively concentrated in the inner segment (where it is found in cells of the intact retina). This phenomenon occurs in cells developing in the absence of intercellular contacts. The development of ATPase polarity correlates with a decrease in the fraction of ATPase molecules that are mobile in the membrane (as determined by fluorescence photobleaching recovery), as well as with an increase in the fraction of ATPase that remains associated with the cells after detergent extraction. The magnitudes of the mobile ATPase fractions agree well with those of the detergent-extractable fractions in both the immature and developed photoreceptors. The distribution of alpha spectrin and ATPase-immunoreactive materials appeared qualitatively similar, and quantitative image analysis showed similar gradients of spectrin and Na+,K+-ATPase immunofluorescence along the long axis of elongated photoreceptors. Moreover, detergent extractability of alpha spectrin and the ATPase showed similar modifications in response to changes in pH or KCl concentration. ATPase detergent-extractable and mobile fractions were not changed in cultures treated with cytoskeletal inhibitors such as nocodazole. These data are consistent with a role for an asymmetrically distributed, spectrin-containing subcortical cytoskeleton in the preferential accumulation of Na+,K+-ATPase in the photoreceptor inner segment.

We have studied the effect of cell density on the lateral diffusion of major histocompatibility (MHC) antigens in the plasma membranes of fibroblasts using fluorescence recovery after photobleaching. The percent recovery of fluorescence was decreased in fibroblasts grown in confluent cultures. While recovery of fluorescence was measurable in greater than 90% of the cells from sparse cultures, measurable recovery was detected in only 60-80% of the cells from dense cultures; no mobile antigens were detectable in 20-40% of cells examined. The diffusion coefficient on human skin fibroblast cells that did show recovery was the same for cells grown in sparse or dense conditions. In WI-38, VA-2, and c1 1d cultures the diffusion coefficients of mobile antigens were smaller in cells from dense cultures. Changes in lateral diffusion occurred with increased cell-cell contact and with age of cell culture but were not observed in growth-arrested cells or in sparse cells cultured in medium conditioned by confluent cells. Decreased mobile fractions of MHC antigens were observed when cells were plated on extracellular matrix materials derived from confluent cultures. Treatment of the extracellular matrix materials with a combination of proteolytic enzymes or by enzymes that degrade proteoglycans abolished this effect. Matrices produced by cells from other cell lines were less effective in inducing changes in mobile fractions and purified matrix components alone did not induce changes in lateral diffusion. Finally, there were no differences in the proportion of MHC antigens that were resistant to Triton X-100 extraction in sparse and dense cells. These results suggest that cell-cell interactions mediated through extracellular matrix materials can influence the lateral diffusion of at least part of the population of MHC antigens.

We have studied the role of restrictions to lateral mobility in the segregation of proteins to apical and basolateral domains of MDCK epithelial cells. Radioimmunoassay and semiquantitative video analysis of immunofluorescence on frozen sections showed that one apical and three basolateral glycoproteins, defined by monoclonal antibodies and binding of beta-2-microglobulin, were incompletely extracted with 0.5% Triton X-100 in a buffer that preserves the cortical cytoskeleton (Fey, E. G., K. M. Wan, and S. Penman. 1984. J. Cell Biol. 98:1973-1984; Nelson, W. T. and P. J. Veshnock. 1986. J. Cell Biol. 103:1751-1766). The marker proteins were preferentially extracted from the "incorrect" domain (i.e., the apical domain for a basolateral marker), indicating that the cytoskeletal anchoring was most effective on the "correct" domain. The two basolateral markers were unpolarized and almost completely extractable in cells prevented from establishing cell-cell contacts by incubation in low Ca++ medium, while an apical marker was only extracted from the basal surface under the same conditions. Procedures were developed to apply fluorescent probes to either the apical or the basolateral surface of live cells grown on native collagen gels. Fluorescence recovery after photobleaching of predominantly basolateral antigens showed a large percent of cells (28-52%) with no recoverable fluorescence on the basal domain but normal fluorescence recovery on the apical surface of most cells (92-100%). Diffusion coefficients in cells with normal fluorescence recovery were in the order of 1.1 x 10(-9) cm2/s in the apical domain and 0.6-0.9 x 10(-9) cm2/s in the basal surface, but the difference was not significant. The data from both techniques indicate (a) the existence of mobile and immobile protein fractions in both plasma membrane domains, and (b) that linkage to a domain specific submembrane cytoskeleton plays an important role in the maintenance of epithelial cell surface polarity.

When the content of cyclic AMP (cAMP) was compared in livers of a series of congenic mouse strains differing at the H-2 locus, significant variation in concentration of cAMP per unit wet weight was found among strains, and also for animals of a given strain with increasing age. For a given age, from 8 to 22 weeks, cAMP levels in liver of H-2a and H-2b genotype animals were significantly higher than that in liver of H-2k type animals. This difference was seen whether the H-2 gene was on the genetic background of strain C57BL/10, C3H, or A. Levels of cAMP in livers of H-2d animals were between those of H-2a and H-2k animals.

In mouse-human heterokaryons, the lateral diffusion of major histocompatibility (MHC) antigens in the plasma membrane is enhanced by treatment of parent cells with ouabain. Ouabain treatment is ineffective if the medium lacks calcium ion, or if Verapamil, a blocker of calcium channels, is present. The divalent ionophore A23187 also enhances lateral diffusion of MHC antigens, to the same extent as ouabain, A23187 is effective only if calcium is present in the medium. Thus it appears that increased levels of cell calcium release constraints to lateral diffusion of MHC antigens.

We have used fluorescence photobleaching and recovery (FPR) to measure the lateral diffusion of mouse H-2 antigens, labeled with fluorescent Fab fragments, in the membrane of cl 1d fibroblasts. Diffusion coefficients, D, vary more than 20-fold from cell to cell, though they vary no more than twofold when measured at different points on a single cell. The fraction of H-2 antigens mobile, R, also varies from cell to cell, and no lateral diffusion of H-2 antigens can be detected in approximately 20% of the cells examined. Treatment of cells with NaCN + NaF, reducing their levels of ATP reduces the proportion of cells in which no lateral diffusion can be detected. The maximum values of D seen in poisoned cells are less than those in controls. Treatment of cells with the divalent inophore, A23187, greatly increases the proportion of cells in which diffusion of H-2 is rapid, D greater than 2 x 10(-9) cm2 s-1. The data obtained on diffusion by FPR can be replotted in the form of an experiment in which lateral diffusion of H-2 antigens is measured in a population of heterokaryons. There is good agreement between this transformation and actual data on heterokaryons. Thus the two methods appear to measure the same transport process.

Rabbit and mouse antisera prepared against teratocarcinoma cells precipitate both glycoproteins and glycolipids from detergent extracts of radiolabeled cells. Extracts of immunoprecipitates with chloroform/methanol, 2:1 (vol/vol) have been resolved on thin-layer gels into multiple peaks. There are more species seen in extracts of teratocarcinoma cells than in extracts of the crossreacting cultured cell line, cl 1d. The teratocarcinoma antigens may be extracted out of chloroform/methanol into buffered saline. Incubation in these secondary extracts converts unreactive cells (lymphocytes to cells reactive with antisera against teratocarcinoma. Furthermore, the coated cells absorb at least 80% of the activity of antisera against teratocarcinoma targets.

Site-directed mutagenesis in vitro was used to introduce stop codons in the genomic DNA of the alpha and beta chains of the murine class II major histocompatibility complex antigen, I-Ak. Mutated DNA was transfected into B lymphoma cells that were then selected by neomycin resistance and for their ability to express I-Ak molecules on their plasma membrane. The translational diffusion coefficient (Dlat) of I-Ak molecules composed of a wild-type beta chain paired with an alpha chain missing either 6 or 12 amino acids from the cytoplasmic domain is on the average threefold higher than the Dlat of wild-type I-Ak molecules as measured by fluorescence photobleaching and recovery. The removal of 12 amino acids from the cytoplasmic domain of the beta chain did not change the Dlat value from that of wild-type I-Ak if the truncated beta chain was paired with a wild-type alpha chain. Removing all amino acids of the cytoplasmic domains of both the alpha and beta chains resulted in a 10-fold increase in the Dlat, the highest value for any of the truncated I-Ak molecules tested. These data indicate that the carboxy-terminal six amino acids of the cytoplasmic domain of the alpha chain and the six plasma membrane-proximal amino acids of the beta chain are important in constraining the translational diffusion of I-Ak molecules in the plasma membrane.

The rate of appearance, in a population of mouse-human heterokaryons, of cells with intermixed mouse and human surface antigens may be used to estimate the rate of lateral diffusion of the antigens in a single cell. Most heterokaryons appear to restrict diffusion of their surface antigens. These restrictions are altered by exposing either heterokaryons or their parent cells to conditions that change cell surface membrane potential. Media containing unphysiological concentrations of potassium ion, drugs, affecting the Na+,K+ ATPase, or a channel-forming antibiotic, gramicidin, all affect lateral mobility of cell surface antigens in a manner consistent with a common effect on membrane potential.

The rate of appearance, in a newly formed heterokaryon population, of cells bearing completely intermixed mouse and human surface antigens may be used to estimate diffusion constants for antigens on individual cells. From this estimate, it appears that the surface antigens in most cells do not diffuse at the rate expected, but rather move more slowly, by a factor of ten or more, than expected from either measured or calculated diffusion constants for proteins freely mobile in the plane of a lipid membrane. Differences in diffusion rates between cells are not due to effects of Sendai virus, or of trypsin. Restrictions on diffusion are apparently not due to cytochalasin B- or Colcemid-sensitive elements.

The rate of collection of embryo fibroblast single cells by an embryo fibroblast monlayer was realted to the H-2 haplotype of the fibroblast monolayer. The rate was highest for the H-2s strains and lowest for the H-2k strains with all other strains examined being intermediate. As opposed to monolayers prepared from the A and C3H background animals, monolayers from B10 background mice only demonstrated an H-2 haplotype dependent rate differential after treatment with fetal calf serum or neuraminidase. The relationship that was seen between monolayer H-2 haplotype and rate of adhesion with embryonic monolayers was not observed with either congenic 3T3 cell lines or fibroblasts derived from adult tissues. It was further shown that the rate of single cell pick-up could be substantially reduced by incubating the monolayers with the appropriate polyspecific anti-H-2 antisera. The inhibition observed appeared to be directly related to anti-H-2 antibody binding and was not merely a function of ligand binding to the cell surface, as antisera directed against other fibroblast cell surface antigens did not significantly inhibit the adhesive rate. These results indicate a role for the H-2 gene complex in modulating fibroblast-fibroblast intercellular adhesion.

Single mouse intestinal epithelial cells (IEC) may be isolated by the use of a combination of methods used for the isolation of IEC from other species. Isolated cells remain viable for several hours. The membrane integral enzymes alkaline phosphatase and leucine aminopeptidase of isolated IEC are localized to the brush borders of IEC in tissue and in most newly isolated IEC. With time, both enzymes are found distributed over the entire cell surface. Redistribution appears to occur by diffusion in the plane of the membrane. It is slowed, but not blocked, if cells are maintained at 0 degrees C instead of at 37 degrees C, and it is not blocked by fixation in 0.5-3% paraformaldehyde. Drugs that alter cell membrane potential or that affect cell levels of ATP enhance the rate of redistribution of the enzymes.

A variety of genetic and biochemical evidence points to an association between major histocompatibility complex (MHC) haplotype and several types of cell surface receptors including epidermal growth factor and insulin receptors. We report evidence for such associations between human class I MHC antigens, HLA antigens, and specific insulin binding sites on human B lymphoblasts. We have measured insulin binding to cells of an HLA-heterozygous, Epstein-Barr virus-transformed B-cell line, LCL 721, and to derivative mutants from which all or part of the HLA complex had been deleted. The affinity, Ka, of insulin binding sites is approximately 10(8) M-1 in mutants expressing antigen HLA-B5 together with other HLA antigens and in mutants expressing only HLA-C. HLA-A1; HLA-A1,B8; HLA-A2,C; and HLA null mutants (not expressing any HLA antigens) bind insulin to sites with an affinity of approximately 10(9) M-1.

We have used the technique of fluorescence photobleaching recovery to measure the lateral diffusion coefficients and the mobile fractions of a fluorescent lipid probe, 1-acyl-2-(12-[(7-nitro-2-1, 3-benzoxadiazol-4-yl)aminododecanoyl]) phosphatidylcholine (NBD-PC), and of labeled membrane proteins of human fibroblasts. Values for mobile fractions decrease monotonically with increasing size of the laser spot used for the measurements, over a range of 0.35-5.0 microns. Values for NBD-PC diffusion coefficients increase in part of this range to reach a plateau at larger laser spots. This variation is not an artifact of the measuring system, since the effects are not seen if diffusion of the probe is measured in liposomes. We also find that the distribution of diffusion coefficients measured with small laser spots is heterogeneous indicating that these small spots can sample different regions of the membrane. These regions appear to differ in protein concentration. Our data strongly indicate that fibroblast surface membranes consist of protein-rich domains approximately 1 micron in diameter, embedded in a relatively protein-poor lipid continuum. These features appear in photographs of labeled cell surfaces illuminated by the expanded laser beam.

The binding of epidermal growth factor (EGF) by A-431 human epidermoid carcinoma cells was reduced after exposure of the cells to low concentrations (0.01-1 mM) of ATP and other nucleoside 5'-triphosphates at 37 degrees C, but not at 0 degree C. This was due to loss of high-affinity EGF binding sites. The modulation was associated with transient increases in inositol phosphate synthesis and intracellular Ca2+ and with phosphorylation of the EGF receptor on serine and threonine. There was no evidence for entry of labeled ATP into the cells. ATP appeared to bind to specific cell surface receptors. Such binding was demonstrated directly with the nonmetabolizable ATP analogue adenosine 5'-[beta,gamma-imido]triphosphate.

Cells from established tissue culture lines of mouse (CIID) and human (VA-2) origin were fused together with Sendai virus, producing heterokaryons bearing both mouse and human surface antigens which were then followed by the indirect fluorescent antibody method. Within 40 min following fusion, total mixing of both parental antigens occurred in over 90 % of the heterokaryons. Mouse H-2 (histocompatibility) and human surface antigens were visualized by successive treatment of the heterokaryons with a mixture of mouse alloantiserum and rabbit anti-VA-2 antiserum, followed by a mixture of fluorescein-labelled goat anti-mouse IgG and tetramethyl- rhodamine-labelled goat anti-rabbit IgG(Fc). The CIID × VA-2 fusions were carried out in suspension and maintained at 37°C in a shaking water bath; aliquots were removed at various intervals and stained with the above reagents. The heterokaryon population was observed to change from an initial one (5-min post-fusion) of non-mosaics (unmixed cell surfaces of red and green fluorescence) to one of over 90 % mosaics (total intermixing of the 2 fluorochromes) by 40 min after fusion. Mouse-human hybrid lines, derived from similar fusions, gave fluorescence patterns identical to those of the mosaic heterokaryons. Four possible mechanisms would yield such results: (i) a very rapid metabolic turnover of the antigens; (ii) integration of units into the membrane from a cytoplasmic precursor pool; (iii) movement, or ‘diffusion ‘of antigen in the plane of the membrane; or (iv) movement of existing antigen from one membrane site into the cytoplasm and its emergence at a new position on the membrane. In an effort to distinguish among these possibilities, the following inhibitor treatments were carried out: (1) both short- and long-term (6-h pre-treatment) inhibition of protein synthesis by puromycin, cycloheximide, and chloramphenicol; (2) short-term inhibition of ATP formation by dinitrophenol (DNP) and NaF; (3) short- and long-term inhibition of glutamine- dependent pathways with the glutamine analogue 6-diazo-5-oxonorleucine; and (4) general metabolic suppression by lowered temperature. The only treatment found effective in preventing the mosaicism was lowered temperature, from which resulted a sigmoidal curve for per cent mosaics versus incubation temperature. These results would be consistent with mechanisms iii and/or iv but appear to rule out i and ii. From the speed with which the antigen markers can be seen to propagate across the cell membrane, and from the fact that the treatment of parent cells with a variety of metabolic inhibitors does not inhibit antigen spreading, it appears that the cell surface of heterokaryons is not a rigid structure, but is ‘fluid’ enough to allow free ‘diffusion’ of surface antigens resulting in their intermingling within minutes after the initiation of fusion.