Cytotoxic T lymphocyte recognition of HLA-A/B antigens introduced into EL4 cells by cell-liposome fusion.

HLA-A2 and -B7 antigens were introduced into EL4 (H-2b) cells by cell-liposome fusion and were used as targets or stimulators for cytotoxic T lymphocytes (CTL) generated in C57B1/6 (H-2b) mice. It was found that such EL4-HLA cells were not recognized by CTL that had been raised against either a human cell line bearing these HLA antigens or the purified HLA-A2 and -B7 antigens reconstituted into liposomes. In addition, EL4-HLA cells were not capable of inducing CTL that could recognize a human cell line bearing HLA-A2 and -B7 antigens. Instead, EL4-HLA cells induced CTL that specifically lysed EL4-HLA cells and not human cells expressing HLA-A2 and -B7. CTL recognition required the presence of HLA antigens on the EL4 cell surface and was inhibited by antibodies against either H-2b or HLA-A/B. Monoclonal antibody binding studies showed that the expected polymorphic determinants of the HLA-A2 and -B7 antigens were still present on EL4-HLA cells. However, the specificity of CTL or their precursors that are capable of recognizing HLA-A2 or -B7 was altered after these antigens became associated with the EL4 surface. Possible explanations for these results are discussed.

Introduction of HLA-A/B antigens into lymphoid cell membranes by cell-liposome fusion.

Conditions were optimized for surface binding by mouse lymphblastoid cell lines of liposomes containing purified human histocompatibility antigens. Antigens were reconstituted into lipid vesicles by detergent dialysis, and were incubated with the acceptor cell lines. After washing, the amount of HLA antigen on the cell surface was quantitated by double antibody radioimmunoassay. Uptake and surface expression was shown to be dependent upon vesicle phospholipid composition, vesicle:cell ratio, acceptor cell line, and stage of cell growth. These studies indicate that by using vesicles composed of 50% phosphatidylethanolamine/50% phosphatidylserine it is possible to introduce into the mouse EL4 cell surface about 70% of the amount of HLA expressed normally on the human lymphoblastoid line JY. The antigen is stably expressed on the surface for several hours and appears to be integrated into the cell membrane, as assessed by both fluorescence microscopy and susceptibility to lysis by using anti-HLA antibody and complement. The method used has the advantage over previously described procedures of not requiring the use of fusogenic proteins or agents such as polyethylene glycol or lysophospholipids, which might perturb overall membrane structure or properties.

Glucocerebroside transfer between phosphatidylcholine bilayers.

We have studied the kinetics of transfer of glucocerebroside between phospholipid bilayers by using pyrene and 3H-labeled glucocerebroside incorporated into dimyristoylphosphatidylcholine (DMPC) and dipalmitoylphosphatidylcholine (DPPC) bilayers. Pyrene-labeled glucocerebroside (PyrCer) molecules are able to form an excited complex (eximer, E) between a PyrCer in the ground state and an excited monomer (M). When vesicles contained a known amount of PyrCer (donors) are incubated with unlabeled vesicles (acceptors), transfer of PyrCer from donor to acceptor populations is reflected in a decrease of the observed E/M intensity ratio. The results obtained from these studies show that the half-time of transfer from donor DMPC-PyrCer vesicles to acceptor DMPC vesicles is greater than 30 days at 37 degrees C. This very slow transfer of glucocerebroside was confirmed by using tritiated glucocerebroside incorporated into small unilamellar DPPC donor vesicles incubated with large unilamellar DPPC acceptor vesicles above the phase transition. Separation of the two vesicle populations by molecular sieve chromatography at 45 degrees C shows a half-time for transfer of approximately 32 days. We conclude that, in contrast to the results obtained for phosphatidylcholines [Roseman, M., & Thompson, T. E. (1980) Biochemistry 19, 439], glucocerebroside does not rapidly transfer between bilayers under these conditions.

Thermotropic behavior of monoglucocerebroside--dipalmitoylphosphatidylcholine multilamellar liposomes.

The thermotropic behavior of multilamellar liposomes prepared from mixtures of glucocerebroside and dipalmitoylphosphatidylcholine has been studied by high-sensitivity scanning calorimetry. It is shown that glucocerebroside has a marked effect on the gel--liquid crystalline transition of dipalmitoylphosphatidylcholine. The pretransition seen in pure samples of dipalmitoylphosphatidylcholine is undetectable at small mode fractions of glucocerebrosides (less than 10%). The main transition is shifted to higher temperatures and becomes broader and less cooperative in the presence of glucocerebroside. The enthalpy change of the main transition decreases with increasing the glucocerebroside content. However, this decrease is not linear with the glucocerebroside/phospholipid mole ratio. Glucocerebroside itself does not show a separate transition in the temperature range of these studies (10--75 degree C). The origin of these effects and their dependence on the glucocerebroside content suggest that the in-plane distribution of glucocerebroside molecules is affected by the physical state of the lipid bilayer and by the glucocerebroside/phospholipid mole ratio.