Many applications envisioned for liposomes in cell biology and chemotherapy require their direction to specific cellular targets. The ability to use antibody as a means of conferring specificity to liposomes would markedly increase their usefulness. We report here a method for covalently coupling soluble proteins, including monoclonal antibody and Staphylococcus aureus protein A (ref. 4), to small sonicated liposomes, by using the heterobifunctional cross-linking reagent N-hydroxysuccinimidyl 3-(2-pyridyldithio)propionate (SPDP, Pharmacia). Liposomes bearing covalently coupled mouse monoclonal antibody against human beta 2-microglobulin [antibody B1.1G6 (IgG2a, kappa) (B. Malissen et al., in preparation)] bound specifically to human, but not to mouse cells. Liposomes bearing protein A became bound to human cells previously incubated with the B1.1G6 antibody, but not to cells incubated without antibody. The coupling method results in efficient binding of protein to the liposomes without aggregation and without denaturation of the coupled ligand; at least 60% of liposomes bound functional protein. Further, liposomes did not leak encapsulated carboxyfluorescein (CF) as a consequence of the reaction.

The in vitro binding of anti-HL-A antibodies t o the membrane of human lymphocytes induces important changes in the distribution of HL-A antigens on the cell surface. Following either direct or indirect immunofluorescence staining at 0 OC, cell-bound anti-HL-A antibodies are dispersed all over the cell surface. When the washed, stained lymphocytes are warmed and incubated at 37 "C, fluorescent antibodies cluster progressively a t the cell surface. They form large spots of fluorescence, and sometimes single caps at one pole of the cell, outside which HL-A antigens are no longer detectable, but other antigens can still b e found. Similar findings were made in electron microscopy, following indirect labeling of HL-A antigens on human lymphocytes, with ferritin or plant virus as the markers.With the indirect immunofluorescence technique, anti-human thymocyte antibodies and anti-mouse H-2 antibodies were found t o induce a similar redistribution of the corresponding antigens.The mechanism and the interpretation of this displacement of surface antigens by antibodies are still unclear and are discussed in terms of membrane structure and immunological significance.

In the preceding papers (1, 2) it was demonstrated that guinea pigs immunized with single proteins or protein-hapten conjugates produced two major types of antibodies differing in their electrophoretic mobility. These antibodies were identified as 71 and 7~; both had a sedimentation coefficient of approximately 7S. Gamma-1 or "fast" migrating antibodies conferred passive cutaneous and systemic anaphylactic reactions in guinea pigs; gamma-2 or "slow" migrating antibodies did not mediate these activities.The present report is concerned with the ability of slow and fast guinea pig antibodies to fix complement in ~tro in the presence of antigen and to participate in certain in vivo reactions, which apparently involve the fixation of complement. It has been found that 7, antibodies bind complement in the presence of antigen and that 71 antibodies fail to do so. These differences may depend on the presence of a complement-fixing site on piece III of 7~ antibodies and the absence of this site from piece III of 71 antibodies (3). The slow antibody components of guinea pig antisera were found to be very effective in conferring passive Arthus reactivity (in the guinea pig), whereas fast components were relatively ineffective. These data suggest that complement activity is involved in the mechanism of the Arthus reaction.Studies of guinea pig anti-sheep erythrocyte sera demonstrated, in addition to the usual slow and fast antibodies, a highly efficient hemolysin of intermediate electrophoretic mobility. Whether this activity is a function of a third type of antibody, which is perhaps produced in response to the particulate properties of the antigen, or to antigenic heterogeneity of the red cell membrane, remains to be determined.