Incorporation into phospholipid vesicles of pore-like properties from Golgi membranes of lactating-rat mammary gland.

The ability of rat mammary-gland Golgi membranes to produce monosaccharide-specific pores in phospholipid vesicles was investigated. The apparent ability of Triton X-100 extracts of Golgi membranes to form such pores was re-evaluated, since we have now found that an apparent pore is produced by the detergent alone. We therefore incorporated intact Golgi membranes (1 mg of protein) into egg-yolk phospholipid vesicles by direct sonication in the absence of any detergent. These vesicles retained about 0.6% of the total sucrose, but demonstrated selective permeability towards glucose compared with sucrose, with 19.8% of the glucose being lost during gel filtration on Sepharose 4B. This phenomenon seemed to be enhanced by the presence of acidic phospholipids and lysophosphatidylcholine, but was inhibited by inclusion of cholesterol in the vesicles. The best mixture of phospholipids comprised 6.5 mg of egg-yolk phospholipid, 1 mg of phosphatidylserine and 0.05 mg of lysophosphatidylcholine, where 32.9% of the glucose was lost. By using this optimum phospholipid mixture the pores were shown to be permeable to both glucose and mannitol, whereas sucrose and lactose were retained by the vesicles. Chaps (3- [(3-cholamidopropyl)dimethylammonio] propane-1-sulphonate)-solubilized membranes produced similar permeability in vesicles produced by dialysis of a solution of the phospholipids mixed with the membrane extract. This technique resulted in a greater loss of glucose, 33% loss requiring about 0.25 mg of protein. The pore-forming ability of both intact Golgi membranes and Chaps extracts was sensitive to boiling and proteolysis, indicating that a membrane protein was likely to be involved in pore formation.