Spin-labeled phospholipids have been used to study the outside -. inside and inside -* outside transport of phospholipids across the human erythrocyte membrane at 3TC. As already shown, inward transport is much faster for aminophospholipids than for phosphatidycholine. In addition, we show here that outward transport of the phosphatidylserine and phosphatidylethanolamine analogues is three to four times faster than that ofphosphatidycholine. Magnesium depletion of the erythrocytes considerably decreases the outward rate of both aminophospholipids to values close to that of phosphatidylcholine. These results suggest that the outward aminophospholipid translocation is, at least partly, protein mediated. The protein involved could be identical to the inward Mg-ATPdependent aminophospholipid carrier.Inward translocation of phosphatidylserine and phosphatidylethanolamine across human erythrocyte membrane is mediated by a specific ATP-dependent protein (1)(2)(3)(4)(5). This selective and active transport has been demonstrated also in other cells (6)(7)(8). To explain the observed phospholipid transbilayer asymmetry (9) in terms of a dynamic equilibrium, existence of an outward motion was postulated (1). Thus, the rate of the outward transport of a given phospholipid can be estimated from knowing both its rate of inward translocation and its asymptotic distribution. Such a calculation was made by Herrmann and Muller (1986) (10) and more recently by Williamson et al. (11).In the present paper we give direct experimental evidence for outward phospholipid motion across the human erythrocyte membrane, and we assess this diffusion rate by using spin-labeled analogues of phospholipids. This outward motion of aminophospholipids is shown to be more than passive diffusion through the bilayer. Evidence for the role of a selective protein carrier is presented.MATERIALS AND METHODS Buffers. The buffer used was 140 mM NaCl/10 mM Hepes, pH 7.4; in some experiments 1 or 2 mM EDTA was added.Erythrocytes. Fresh human blood was obtained from healthy volunteers or from a local blood bank (Hopital Cochin). Blood collected on EDTA was washed three times in 3 vol of buffer by centrifugation (1000 g at 4°C for 5 min). The washed erythrocytes were stored on ice and used within 6 hr. All samples contained 5 mM diisopropyl fluorophosphate (6) to minimize hydrolysis of spin-labeled phospholip- . Spin labels were synthesized as described (1,12,13). Suitable amounts of spin labels in chloroform or chloroform/methanol 2:1 (vol/vol) were dried under vacuum and resuspended in buffer. Two volumes of this aqueous dispersion of labels (ice-cold) were then added to one volume of ice-cold washed erythrocyte pellet. We verified, as in ref.1, that the spin-labeled phospholipids incorporate in erythrocyte membranes within 3 min under such conditions. The spin-label concentration, unless specified, was 2 mol % ofthe endogenous phospholipids. Spontaneous reduction of the spin-labeled concentration was monitored after bovine serum albumin (BSA) extracti...