Coformulated N-Octanoyl-glucosylceramide Improves Cellular Delivery and Cytotoxicity of Liposomal Doxorubicin

The anthracycline doxorubicin has been in clinical use for several decades and is still among the most widely used chemotherapeutic agents for treatment of a variety of neoplasms (Weiss, 1992; Lothstein et al., 2001; Zagotto et al., 2001). Despite many years of research in developing new and better anthracyclines, little or no change in the molecular structure of doxorubicin made it to the clinic. However, with the development of liposomal formulations, its delivery to at least certain types of tumors underwent a major improvement (Tardi et al., 1996; Gabizon, 2001). Compared with systemic application of free doxorubicin, liposomal doxorubicin exhibits significant advantages, such as reduced normal tissue toxicities (Orditura et al., 2004). Improved loading procedures, resulting in high doxorubicin encapsulation efficiencies, further increased the therapeutic index of entrapped doxorubicin (Papahadjopoulos et al., 1991; Horowitz et al., 1992; Haran et al., 1993). Another major step forward was the development of polyethylene glycol (PEG)-coated liposomes. This coating prevents opsonization and reduces the uptake by macrophages from the reticulo-endothelial system, in turn resulting in prolonged circulation times, compared with free doxorubicin or with noncoated liposomes (Papahadjopoulos et al., 1991; Vaage et al., 1992; Robert and Gianni, 1993; Gabizon et al., 1996; Uster et al., 1996). PEG-liposome-encapsulated doxorubicin (commercially available as Caelyx and Doxil) is now in clinical use for the treatment of various types of neoplasms, and innovations such as covalent attachment of tumor-specific antibodies or ligands will further enhance drug targeting and therapeutic value (Park et al., 2002; Pan et al., 2003).

The endothelial lining of healthy blood vessels effectively prevents escape of liposomes from the circulation. In contrast, angiogenesis-associated vascular abnormalities of many solid tumors do allow extravasation of long-circulating PEG-liposomes into the tumor stroma (Yuan et al., 1994). Despite the resulting accumulation in tumor tissue, the liposomes are not taken up by the tumor cells. Instead, their doxorubicin load is gradually released into the interstitial space (Horowitz et al., 1992; Harasym et al., 1997). Given the intracellular localization of its molecular targets, sufficient cellular uptake of doxorubicin is required for its action (Speth et al., 1988; Lothstein et al., 2001). However, doxorubicin does not possess the optimal degree of lipophilicity for efficient plasma membrane traversal, which limits its efficacy (Heijn et al., 1999; Washington et al., 2001).

We recently demonstrated that the cellular uptake of free doxorubicin, and with that its cytotoxic action, is greatly enhanced when the short-chain sphingolipid N-hexanoyl-sphingomyelin (C₆-SM) is coadministered with the drug in vitro (Veldman et al., 2004). In vivo, however, doxorubicin and the sphingolipid will most likely show differences in biodistribution, metabolic kinetics and clearance, and will thus not be delivered at the same site at the same time, which is a prerequisite for the drug uptake-enhancing effect.

In the present study, we investigated the in vitro properties and efficacy of PEG-liposomal doxorubicin that was coformulated with N-octanoyl-glucosylceramide (C₈-GlcCer), another doxorubicin uptake-enhancing sphingolipid that we identified. We choose C₈-GlcCer because it exhibits lower cellular toxicity than C₆-SM. Incorporation of this sphingolipid into the liposome bilayer would ensure codelivery of doxorubicin and the lipid, as required. We report here that this adaptation greatly enhances doxorubicin transfer to tumor cells, in turn leading to increased cytotoxicity. Importantly, this was also the case at high serum concentrations, which holds promise for in vivo application.

Materials and Methods

Materials. Dipalmitoylphosphatidylcholine (DPPC) and distearylphosphatidylethanolamine (DSPE)-PEG2000 were from Lipoid (Ludwigshafen, Germany), and C₈-GlcCer was from Avanti Polar Lipids (Alabaster, AL). Polycarbonate filters were from GE Osmonics (Minnetonka, MN), and PD-10 Sephadex columns were from Pfizer, Inc. (New York, NY). Doxorubicin·HCl was obtained from Pharmachemie (Haarlem, The Netherlands), and Caelyx was from Schering-Plough (Heist-op-den-Berg, Belgium). Cholesterol, bicinchoninic acid protein kit, MTT, and Dowex 50WX4-400 resin (Dowex) were from Sigma-Aldrich (St. Louis, MO). LK6D silica TLC plates were from Whatman (Maidstone, UK), and Vectashield mounting medium was from Vector Laboratories (Burlingame, CA). CytoTox 96 lactate dehydrogenase (LDH) activity kit was from Promega (Madison, WI).

Liposome Preparation and Analysis. Liposomes were prepared by lipid film hydration and extrusion and subsequent remote loading using an ammonium sulfate method (Haran et al., 1993). Mixtures of DPPC, cholesterol, DSPE-PEG2000, and C₈-GlcCer were prepared in various molar ratios (Table 1). Lipids dissolved in ethanol were mixed, and a lipid film was created under reduced pressure on a rotary evaporator and then dried under a stream of nitrogen. Liposomes were then formed by addition of a 250 mM (NH₄)₂SO₄ solution to the lipid film. The hydrated lipid dispersion was sized, successively, by extrusion through 200-nm (three passes), 100-nm (three passes), and double 50-nm (six passes) polycarbonate filters. This standardized procedure combines easy production (at low pressure), high-lipid recovery, and the production of a homogeneous population of small (100-nm) unilamellar vesicles. Nonencapsulated (NH₄)₂SO₄ was removed by gel permeation chromatography using a PD-10 Sephadex column (Pfizer, Inc.), eluted with 123 mM citrate buffer, pH 5.5. For liposomal doxorubicin loading, the drug was added in a molar ratio of doxorubicin to total lipid of 1 to 5 and incubated for 1 h at 55°C (Haran et al., 1993). Nonencapsulated doxorubicin was removed by gel permeation chromatography using a PD-10 Sephadex column, eluted with 135 mM NaCl in 10 mM HEPES, pH 7.4. This entire procedure was used for all liposome formulations.

For postinsertion of C₈-GlcCer into Caelyx, the liposomes were incubated (30 min at 37°C) with various concentrations of the sphingolipid (Fig. 5), which was added as an ethanolic solution [never exceeding 2% (v/v)]. Liposomes were then subjected to ultracentrifugation in a Kontron TFT 80.2 rotor (70,000 rpm, 1 h at 20°C), and pellets were washed with 135 mM NaCl in 10 mM HEPES, pH 7.4.

Phospholipid phosphorus of each liposome preparation was determined by a phosphate assay after perchloric acid destruction (Rouser et al., 1970). To confirm C₈-GlcCer incorporation and postinsertion, lipids were extracted from the liposomes (Bligh and Dyer, 1959). Extracts were then applied to a 60-Å silica gel TLC plate, which was developed in chloroform/methanol/water [60:30:8 (v/v/v)]. Lipids were visualized by iodine vapor staining, and C₈-GlcCer was identified by corunning a standard on the same plate. Liposomal doxorubicin was measured by fluorescence after solubilization of the liposomes in 1% (w/v) Triton X-100, by comparison with standard amounts. Particle size and size distribution were determined by dynamic laser light scattering using a Malvern 4700 system equipped with a 75-milliwatt argon laser (Uniphase, San Jose, CA), with data analysis by Automeasure software, version 3.2 (Malvern Ltd., Malvern, UK). The mean diameter of the liposomes was obtained from the volume distribution curves produced by the particle analyzer. The liposomes used in the present study had a mean particle size of 0.1 μm and a polydispersity <0.1. Polydispersity is a measure for variation in particle size within a liposome population and varies from 0 (complete monodispersity) to 1 (large variations in particle size). The extent of doxorubicin leakage was measured after storing the liposomes for various periods and at various temperatures (Table 1). Samples were then subjected to ultracentrifugation as described above. Released doxorubicin was measured fluorometrically in aliquots of the supernatants using a PerkinElmer Victor Wallac II fluorescence plate reader. Excitation wavelength was 485 nm, and emission wavelength was 535 nm. Background fluorescence was subtracted, and doxorubicin leakage was expressed relative to the total liposomal fluorescence measured after solubilization in 1% Triton X-100.

Cell Culture. A431 cells (adherent human epidermoid carcinoma cells), purchased from the American Type Culture Collection (Manassas, VA), were cultured in Dulbecco's modified Eagle medium, supplemented with 10% (v/v) fetal calf serum, 100 units/ml penicillin, 100 μg/ml streptomycin, and 4 mM l-glutamine. Cells were subcultured once a week by trypsinization and maintained in a water-saturated atmosphere containing 5% CO₂ at 37°C. All experiments were performed at a confluence of 80 to 90%.

Cellular Doxorubicin Uptake. A431 cells were cultured in flat-bottom 96-well plates. At confluence, cells were changed to serum-free medium and exposed to 50 μM free or to 20 μM liposomal doxorubicin. These concentrations were chosen because they closely match the actual plasma concentrations that are reached upon intravenous bolus administrations into average-sized persons [note that clinically used standard dosages differ for free doxorubicin and liposomal doxorubicin (50-75 and 20-40 mg/m², respectively)]. After extensive washing, cells were lysed in 100 μl of 1% (w/v) Triton X-100. Doxorubicin fluorescence was then measured by a PerkinElmer Victor Wallac II fluorescence microplate reader using 485- and 535-nm filters for excitation and emission, respectively. All values were corrected for background fluorescence. Cellular doxorubicin contents were calculated with the aid of standard amounts and corrected for any differences in protein content, as determined with the bicinchoninic acid assay (Smith et al., 1985).

Fluorescence Microscopy. For microscopic studies, A431 cells were cultured on 0.5% (w/v) gelatin-coated glass coverslips. After exposure to liposomal doxorubicin, cells were washed, fixed [10 min in 4% (w/v) paraformaldehyde in phosphate-buffered saline], and mounted in Vectashield on glass slides. All samples were examined with a Zeiss Axiovert S 100 inverted fluorescence microscope using a mercury lamp in combination with a filter set consisting of a 450- to 490-nm band pass excitation filter, a 510-nm beam splitter, and a 520-nm-long pass emission filter. Specimens were photographed through a 40× oil immersion objective by a Zeiss AxioCam charge-coupled device camera using equal exposure times (1350 ms).

Plasma Membrane Leakage Assay. To investigate (loss of) plasma membrane integrity, the release of the cytosolic enzyme LDH into phenol red-free cell culture medium was monitored. Flat-bottom 96-well plates containing (liposome-treated) A431 cells were centrifuged (2000 rpm, 5 min at 20°C), and 50-μl aliquots of the supernatants were assessed for LDH activity with the CytoTox 96 kit according to the manufacturer's instructions. Values obtained from cells that were lysed in 0.8% Triton X-100 served as positive control. All data were corrected for protein contents, as determined with the bicinchoninic acid assay (Smith et al., 1985).

Cell Viability. For viability assessment, A431 cells were cultured in flat-bottom 96-well plates. After experimental treatments, 100 μg of the mitochondrial dehydrogenase substrate MTT was added to each well (Carmichael et al., 1987). Cells were then incubated for 60 min at 37°C. After centrifugation (3000 rpm, 10 min at 4°C) supernatants were removed. The precipitated blue formazan products were then dissolved in 100 μl of dimethyl sulfoxide, and absorbencies were read in a Bio-Tek Instruments EL 340 photospectrometric microplate reader at 540 nm (Bio-Tek Instruments, Winooski, VT). Background absorbencies were subtracted, and values from wells containing untreated control cells were set at 100% viability.

Results

C₆-SM Enhances Cellular Doxorubicin Uptake but Is Cytotoxic at High Concentrations, Whereas C₈-GlcCer Shows the Same Uptake Enhancement but Is Less Toxic. We previously reported that the cellular uptake of free doxorubicin could be greatly (more than 3-fold) enhanced in vitro by coaddition of C₆-SM, a semisynthetic short-chain sphingolipid (Veldman et al., 2004). At its effective low micromolar concentrations and during short-term incubations (serum-free), C₆-SM displayed no toxic effects by itself. However, higher concentrations (Fig. 1) and/or longer incubation periods (data not shown) with this lipid were clearly toxic. Overall, viability of A431 cells dramatically decreased at C₆-SM concentrations that exceed 10 μM (Fig. 1A). This is partly the result of loss of plasma membrane integrity by intrusion of this lipid, as demonstrated by the dose-dependent leakage of the cytosolic enzyme LDH into the extracellular medium (Fig. 1B).

In a search for less toxic alternatives, we now identified C₈-GlcCer as another sphingolipid with equally potent doxorubicin uptake-enhancing properties (335.8 ± 88.5% at 10 μM in A431 cells) as C₆-SM. In contrast to C₆-SM, however, C₈-GlcCer clearly exhibited less toxicity at high concentrations (Fig. 1A). In addition, C₈-GlcCer caused no leakage of LDH, even at concentrations as high as 100 μM (Fig. 1B), indicating that the plasma membrane remained completely intact.

Characterization of C₈-GlcCer-Enriched PEG-Liposomes. To investigate whether C₈-GlcCer retains its doxorubicin uptake-enhancing properties when coformulated with liposomal doxorubicin, we prepared a series of PEG-liposomes, containing increasing mol% C₈-GlcCer (Table 1). Whereas the relative amount of PEG-containing phospholipid (DSPE-PEG2000) was kept constant, DPPC and cholesterol contents slightly decreased in this series to compensate for increasing mol% C₈-GlcCer. Liposomes were loaded with doxorubicin as described previously (Haran et al., 1993). The lipid composition, doxorubicin loading efficiencies, size, and stability of the liposome preparations fully matched those of commercial PEG-liposomal doxorubicin (Caelyx). Because truncated lipid analogs are known to destabilize the lipid bilayer, we limited the C₈-GlcCer content to 17 mol%. At this C₈-GlcCer content, the liposomes were still very stable, as shown by the minimal doxorubicin leakage (at 37°C or during prolonged storage at 4°C) (Table 1).

Enhanced Cellular Doxorubicin Uptake from C₈-GlcCer-Enriched Liposomes. When cultured A431 cells were incubated for 24 h in the presence of normal (conventional) PEG-liposomal doxorubicin, 2.6 ± 0.1 nmol of drug/mg of cellular protein accumulated in the cells. When using liposomes that were enriched with C₈-GlcCer (up to 17%), however, this uptake increased up to 4-fold (Fig. 2). This observation was confirmed by fluorescence microscopy: Doxorubicin accumulation in cell nuclei during 24-h incubation with C₈-GlcCer-enriched liposomes (Fig. 3C) was much higher than with conventional liposomes (Fig. 3B).

Improved Doxorubicin Delivery to Cells Results in Enhanced Cytotoxicity. The in vitro efficacy of the modified liposomes was tested by the MTT viability assay. This assay relies on the fact that only viable cells have the required mitochondrial metabolic capacity to convert MTT into a detectable product, whereas dead or dying cells have not (Carmichael et al., 1987). Compared with a control treatment (empty liposomes; set at 100%), conventional doxorubicin-loaded PEG liposomes exhibited only modest cytotoxicity (58.4% A431 cells was still viable after 24 h; Fig. 4). At increasing mol% C₈-GlcCer, however, this viability decreased dramatically and correlated well with increasing doxorubicin uptake of the cells (Figs. 2 and 3). The increased potency of the adapted liposomes can be fully attributed to doxorubicin, because control liposomes (without doxorubicin and without C₈-GlcCer; or without doxorubicin, but with C₈-GlcCer) had no significant effect on cell viability. These results were confirmed by viability assessment through phase-contrast microscopy (Fig. 5).

An Alternative Preparation Procedure: Postinsertion of C₈-GlcCer into Caelyx. We tested whether postinsertion of C₈-GlcCer into existing (commercial) doxorubicin liposomes (Caelyx) would also result in improved cytotoxic efficacy. For this, Caelyx was incubated for 30 min in buffer supplemented with various concentrations of C₈-GlcCer. After washing the liposomes, TLC analysis revealed that this simple procedure sufficed to incorporate substantial amounts of C₈-GlcCer into Caelyx in a concentration-dependent manner (Fig. 6, inset). However, Because of the presence of the PEG shielding, it is difficult to assess on theoretical grounds, whether the lipid actually inserted into the liposomal bilayer or merely remained associated to the liposomal surface.

Figure 6 shows that C₈-GlcCer-loaded Caelyx resulted in a similar improved doxorubicin transfer to cells as the newly prepared liposomes (compare with Fig. 2). Accordingly, this was likewise accompanied by an increased cytotoxicity, as determined by the MTT viability assay. Although 76.2 ± 18.2% A431 cells were still viable after 24-h incubation with normal Caelyx, C₈-GlcCer-enriched Caelyx dropped this viability value down to 19.2 ± 11.8%. Because liposome production involves tedious and time-consuming procedures, postinsertion thus provides a convenient alternative for quick preparation of C₈-GlcCer-enriched batches of liposomal doxorubicin, with improved cytotoxicity.

C₈-GlcCer-Enriched Liposomes Are Relatively Insensitive to Serum Components. As described, under serum-free conditions, C₈-GlcCer greatly improves the cellular uptake of free doxorubicin. In the presence of serum, however, both the doxorubicin uptake and the enhancing effect of the sphingolipid hereupon were strongly reduced (Fig. 7A). Most likely, this is due to binding of the amphiphilic doxorubicin and C₈-GlcCer to and sequestration by albumin and lipoproteins, thus preventing interaction of these molecules with cells. Although less pronounced, also the uptake of doxorubicin from conventional PEG-liposomes was diminished by the presence of serum (Fig. 7B). However, it is noteworthy that the relative effect of coformulated C₈-GlcCer was fully retained under high-serum conditions. In the presence of 50% (v/v) serum, for example, 8.6 ± 0.4 nmol/mg doxorubicin was taken up from C₈-GlcCer-enriched liposomes, whereas this was only 1.6 ± 0.3 nmol/mg from an equal amount of C₈-GlcCer-free liposomes. Comparable results were obtained with (postinserted) C₈-GlcCer-Caelyx (data not shown).

As expected, high serum concentrations had a strong protective effect on the viability of doxorubicin-treated A431 tumor cells (not shown in figure). In the presence of 50% serum, for example, 85.4 ± 9.1% of the cells survived a 24-h treatment with conventional liposomal doxorubicin. When using C₈-GlcCer-enriched doxorubicin liposomes, however, only 35.7 ± 6.2% of the cells survived under these conditions. Again, the increased drug uptake correlated well with an increased toxicity. It can be concluded that, under high serum conditions (like in vivo), C₈-GlcCer-enriched doxorubicin liposomes retain higher cytotoxicity than conventional liposomes (Caelyx) or free doxorubicin.

Liposome-Cell Contact Is Required for C₈-GlcCer Action but Not for Doxorubicin Transfer. To address the question whether doxorubicin transfer from liposomes to cells requires the direct interaction of the liposome with the cell surface (plasma membrane), we used Dowex, a nontoxic ion-exchange resin that binds to free doxorubicin with high affinity, but not to liposome-encapsulated doxorubicin (Storm et al., 1985; Druckmann et al., 1989). Supplementation of cell culture media with 5 mg/ml Dowex, which is itself not taken up by cells, prevented death of A431 cells, as induced by a 24-h treatment with conventional PEG-liposomal doxorubicin (90.4 ± 24.9% surviving cells). Likewise, cell death induced by 17 mol% C₈-GlcCer-containing liposomal doxorubicin could also be largely prevented by Dowex cotreatment (81.9 ± 6.3% survival). These findings suggest that doxorubicin is delivered to cells via the aqueous extracellular environment, rather than by direct membrane-membrane contact.

We next questioned how C₈-GlcCer in the liposome improves the cytotoxicity of liposomal doxorubicin. Is there any interaction or complex formation between the lipid and the drug prior to interaction with the cell membrane? To address this question, we preincubated A431 cells with the sphingolipid and subsequently removed nonbound lipid by washing. This treatment sufficed to improve the cellular uptake of subsequently added doxorubicin (data not shown). Likewise, when A431 cells were treated with a combination of free doxorubicin together with empty 17 mol% C₈-GlcCer liposomes, drug uptake increased to 260.4 ± 28.2%, compared with empty liposomes without C₈-GlcCer (set at 100% uptake). In addition, doxorubicin uptake from normal liposomes (0 mol% C₈-GlcCer) improved when C₈-GlcCer was added separately, rather than coformulated in the liposome (data not shown). Together, these results indicate that the sphingolipid acts at the level of the plasma membrane, and not within the liposome or the aqueous environment. Furthermore, the doxorubicin-uptake enhancing effect neither requires simultaneous introduction of C₈-GlcCer and doxorubicin into the plasma membrane nor complex formation between the lipid and the drug before interaction with the cell membrane.

Discussion

In this article, we have shown that the cellular uptake of doxorubicin from C₈-GlcCer-enriched doxorubicin liposomes (either newly prepared or by postinsertion into Caelyx) is superior both to free doxorubicin and to conventional liposomal doxorubicin (normal Caelyx), even under high serum conditions in vitro. These results suggest the feasibility of in vivo applications of C₈-GlcCer-modified liposomes.

We have addressed the mechanism by which liposomal C₈-GlcCer enhances the cellular uptake of doxorubicin. We know that PEG-liposomes, due to their hydrophilic coating, are not taken up by cells and do not fuse with the plasma membrane (Koning et al., 1999, 2003; Everts et al., 2003). Instead, their doxorubicin content is gradually released free into the surrounding medium, from where it is then taken up by neighboring cells (Horowitz et al., 1992; Harasym et al., 1997). Incorporation of C₈-GlcCer into the liposomal bilayer did not lead to enhanced doxorubicin leakage, and neither did it result in improved doxorubicin loading of liposomes (Table 1). Apparently, the sphingolipid does not affect doxorubicin diffusion through the liposomal bilayer, and therefore it does not enhance its release rate. There is no direct liposome-cell contact required for the transfer of doxorubicin to the plasma membrane. This was demonstrated by the prevention of doxorubicin-induced cell death by using the ion-exchanger Dowex that binds to free doxorubicin but not to liposome-encapsulated doxorubicin. These data together with the drug-sequestration effect of serum (Fig. 7B), indicate that doxorubicin is delivered to cells via the aqueous extracellular environment, rather than by direct membrane-membrane contact.

We previously demonstrated that plasma membrane-inserted short-chain sphingolipids, such as C₈-GlcCer, improve the uptake of free (nonliposomal) doxorubicin, by facilitating its trans-membrane diffusion (Veldman et al., 2004). The molecular and cellular mechanisms involved are not fully understood and are currently under investigation. Importantly, simultaneous presence of free drug and the lipid in the culture media was not required for the effect, and no drug-lipid complex formation, before interaction with the cell membrane, was involved. In fact, a preincubation of cells with the sphingolipid (in free or liposomal form) and subsequent removal of nonbound lipid, was sufficient to improve the cellular uptake of subsequently added doxorubicin. These results indicate that the sphingolipid acts at the level of the cell (membrane) and not within the liposome or the aqueous environment.

An important question is how C₈-GlcCer moves from the liposomes to the plasma membrane. Interestingly, serum does not inhibit the relative effect of C₈-GlcCer on liposomal doxorubicin uptake, despite the presence of large amounts of lipid-binding entities (Fig. 7B). This suggests that, contrary to doxorubicin, C₈-GlcCer could (partly) relocate via direct contact between cell-associated liposomes and the plasma membrane. Such a (spontaneous) redistribution, very common for short-chain lipid analogs, would eventually result in an equilibrium in C₈-GlcCer distribution between cells and liposomes (Jeckel and Wieland, 1993; Bai and Pagano, 1997). A similar transfer to tumor cell membranes was described for the liposomal bilayer-incorporated prodrug 5-fluorodeoxyuridine-dipalmitate (Koning et al., 1999).

In summary, we may explain the improved cell delivery of C₈-GlcCer-enriched liposomal doxorubicin by the following sequence of events: Upon addition of the liposomes, doxorubicin is gradually released free into the cell culture medium, just like from normal liposomes. From those liposomes that (transiently) come into physical contact with cells, C₈-GlcCer spontaneously undergoes interbilayer movement toward the plasma membrane, whereas the other liposome components do not. Future studies should confirm that C₈-GlcCer is indeed taken up by the cells, in the same way as C₆-SM (Veldman et al., 2004). We hypothesize that C₈-GlcCer, after lateral diffusion in the plane of the plasma membrane, creates local areas that exhibit an increased permeability toward amphiphilic drugs. This would result in an enhanced cellular influx and accumulation of doxorubicin, in turn explaining the improved cytotoxic action of C₈-GlcCer-enriched liposomes.

Concluding Remarks. The anthracycline doxorubicin has been in widespread clinical use for several decades. Nevertheless, its delivery to tumor cells and therewith its efficacy is far from optimal. Even currently used liposomal formulations do not fully exhibit the desired pharmacological properties. By C₈-GlcCer enrichment, we now introduce a simple but very effective modification, which significantly improves doxorubicin transfer from liposomes into cells, compared with existing formulations. This finding might be a next step forward in the evolution of doxorubicin pharmacology and holds promise for clinical application.