Adriamycin

ABSTRACT A stopped-flow spectrofluorometer equipped with a rapid scanning emission monochromator was utilized to monitor the binding of adriamycin to phospholipid liposomes. The latter process is evident as a decrease in fluorescence emission from a trace amount of a pyrene-labeled phospholipid analog (PPDPG,1 -palmitoyl-2-[(6-pyren-1-yl)]decanoyl-snglycero-3-phospho-rac-glycerol) used as a donor for resonance energy transfer to adriamycin. For zwitterionic 1-palmitoyl2-oleoyl-sn-glycero-3-phosphocholine (POPC) liposomes, fluorescence decay was slow, with a half-time t^sub 1/2^ of -2 s. When the mole fraction of the acidic phospholipid, 1-palmitoyl-2-oleoyl-sn-glycero-3-phospho-rac-glycerol (POPG), was increased to X^sub PG^ >= 0.04, the decay of fluorescence became double exponential, and an additional, significantly faster process with t^sub 1/2^ in the range between 2 and 4 ms was observed. Subsequently, as X^sub PG^ was increased further, the amplitude of the fast process increased, whereas the slower process was attenuated, its t^sub 1/2^ increasing to 20 s. Increasing [NaCI] above 50 mM or [CaCI^sub 2^] above 150 (mu)M abolished the fast component, thus confirming this interaction to be electrostatic. The critical dependence of the fast component on X^sub PG^ allows the use of this process to probe the organization of acidic phospholipids in liposomes. This was demonstrated with 1,2-palmitoyl-sn-glycero-3-phosphocholine (DPPC) liposomes incorporating PPDPG (X^sub PPDPG^ = 0.03), i.e., conditions where X^sub PG^ in fluid bilayers is below the required threshold yielding the fast component. In keeping with the presence of clusters of PPDPG, the fast component was observed for gel-state liposomes. At -34 deg C (i.e., 6 deg below T^sub m^), the slower fluorescence decay also appeared, and it was seen throughout the main phase transition region as well as in the liquid-crystalline state. The fluorescence decay behavior at temperatures below, above, and at the main phase transition temperature is interpreted in terms of thermal density fluctuations and an intermediate state between gel and liquid-crystalline states being involved in the phospholipid main phase transition. This is the first observation of a cluster constituted by acidic phospholipids controlling the membrane association of a drug.

INTRODUCTION

Lateral heterogeneity and domain formation in biomembranes are well established (for reviews see Knoll et al., 1991; Rodgers and Glaser, 1991; Kinnunen, 1991; Gascard et al., 1993; Tocanne et al., 1994; Welti and Glaser, 1994) and have been shown to be of importance to the membrane binding and function of proteins such as cytochrome c (Vanderkooi et al., 1973; Mustonen et al., 1987) and gelsolin (Janmey and Chaponnier, 1995), in the coupling of G-proteins to their receptors (Neubig, 1994), in the budding of viruses from the plasma membrane (Luan and Glaser, 1994), and in the activation of pancreatic lipase by colipase (Momsen et al., 1995). Considering the coupling between dynamic order and function emphasized in the modern view of biological membranes (see e.g., Kinnunen, 1991; Kinnunen and Mouritsen, 1994; Kinnunen et al., 1994; Mouritsen and Kinnunen, 1996), thorough understanding of the mechanisms causing lateral organization of membranes is clearly of fundamental importance. To this end, several molecular-level mechanisms have by now been established in model membranes, as follows.

Characteristically for liquid crystalline materials, phospholipids can exist in several phases such as gel, liquid crystalline, interdigitated, or inverted nonlamellar phases (Kinnunen and Laggner, 1991). Both the coexistence of gel and fluid (liquid crystalline) phases and fluid-fluid immiscibility have been demonstrated for binary alloys (e.g., Shimshick and McConnell, 1973; Wu and McConnell, 1975; Holopainen et al., 1997). Dehydration-induced phase separation was predicted on theoretical grounds (Bryant and Wolfe, 1989) and was recently verified for binary liquid crystalline bilayers (Lehtonen and Kinnunen, 1995). There is evidence of lateral heterogeneity in lipid membranes being induced by alcohol (Rowe, 1987; Mou et al., 1994). In binary membranes containing charged (anionic or cationic) species, phase separation can be induced by Ca2+ (Galla and Sackmann, 1975; Eklund et al., 1988), polycations (Hartmann et al., 1977; Eklund and Kinnunen, 1986), DNA (Koiv et al., 1994), electric fields (Lee et al., 1994; Lee and McConnell, 1995; Soderlund et al., 1997), and pH (Tilcock and Cullis, 1981). Likewise, several cationic peripheral proteins induce the segregation of acidic phospholipids (Kinnunen et al., 1994). In tertiary alloys the cationic amphiphile sphingosine seems to form microdomains with phosphatidic acid (Mustonen et al., 1993). The mismatch between the lengths of the hydrophobic membrane-spanning segments of integral membrane proteins and those of lipids was proposed as a mechanism causing ordering of membranes (Mouritsen and Bloom, 1984). This principle was recently confirmed experimentally for Escherichia coli lactose permease (Lehtonen and Kinnunen, 1997) and bacteriorhodopsin (Dumas et al., 1997) reconstituted into liposomes. Hydrophobic matching conditions have also been shown to induce lipid microdomains in fluid, liquid crystalline phospholipid alloys (Lehtonen el: al., 1996). Steric elastic strain has been suggested to provide the driving force for the formation of well-defined superlattices in liquid crystalline liposomes (for brief reviews see Kinnunen, 1991; Kinnunen et al., 1993).

The biological significance of the rich scale of different lipid phases and phase transitions has remained mostly unknown. Phospholipid chain melting at the main transition occurs via the growth of fluid lipid domains in a bulk gel-state membrane. Accordingly, the main transition readily involves dynamic ordering and lateral heterogeneity in the membrane (Mouritsen, 1991, and references therein), which further have profound effects on membrane properties and on lipid-protein interactions, for instance (for a brief review see, e.g., Mouritsen and Kinnunen, 1996). To this end, although dynamic lateral heterogeneity is indeed inherent to any two-dimensional melting process, it is still extremely challenging to describe the relevant processes in terms of molecular interactions (Jutila and Kinnunen, 1997).

Several drugs bearing a positive charge have been shown to interact with acidic phospholipids. In this respect, perhaps the best characterized is adriamycin (e.g., Karczmar and Tritton, 1979; Goormaghtigh et al.,1980a; Henry et al., 1985; de Wolf et al., 1991), a chemotherapeutic drug that is highly soluble in water yet also effectively partitions into membranes. Drug-lipid interactions are believed to be important to the cytotoxicity of adriamycin and may also underlie its clinically most feared, dose-dependent side effect, cardiotoxicity, which involves impairment of mitochondrial function (e.g., Rosenoff et al., 1975; Solem et al., 1996; Doroshow, 1991; Duarte-Karim et al., 1976; Goormaghtigh et al., 1980b, 1982, 1986; Huart et al., 1984; Nicolay and de Kruijff, 1987). The mechanism of adriamycin cytotoxicity appears to be the induction of apoptosis (e.g., Skladanowski and Konopa, 1993; Ling et al., 1993; Zaleskis et al., 1994; Anand et al., 1995; Miller et al., 1997). In this regard, the activation of phospholipase A2 by adriamycin demonstrated in vitro to result from drug-phospholipid interaction could be relevant (Mustonen and Kinnunen, 1991).

Two types of binding have been suggested to be involved in adriamycin-acidic phospholipid interactions (Karczmar and Tritton, 1979; Henry et al., 1985). In the first mechanism there is an electrostatic association of the amino sugar moiety of the drug with the ionized phosphate of acidic phospholipid, while the dihydroxyanthraquinone ring structures would remain outside the bilayer and may further interact with each other by forming stacks (Goormaghtigh and Ruys-schaert, 1984; Henry et al., 1985; Nicolay et al., 1988). Interaction of the amino sugar with the ionized phosphate is also central in the second mechanism, which additionally involves intercalation of the dihydroxyanthraquinone moiety between the fatty acid chains of phospholipids within the lipid membrane (Dupou-Cezanne et al., 1989; Henry et al., 1985; Fiallo and Garnier-Suillerot, 1986). The above conclusions have been derived from steady-state equilibrium studies (e.g., Karczmar and Tritton, 1979; Goormaghtigh et al., 1980b; Henry et al., 1985; de Wolf et al., 1991).

To our knowledge, the binding of adriamycin to phospholipids has not been assessed by stopped-flow techniques. We report here on the dependence of the rate of the binding of adriamycin to liposomes on their content of the acidic phospholipids. Subsequently, we used the kinetics of drugmembrane interaction to probe membrane lateral order in the course of the main transition of a binary phospholipid alloy.

MATERIALS AND METHODS

Materials

HEPES, EDTA, and adriamycin (doxorubicin) were from Sigma, and NaCl and CaCl^sub 2^ were from Merck (Merck, Darmstadt, Germany). 1-Palmitoyl2-oleoyl-sn-glycero-3-phosphocholine (POPC) and 1-palmitoyl-2-oleoylsn-glycero-3-phosphoglycerol (POM) were from Avanti Polar Lipids (Alabaster, AL). Pyrene-containing phospholipid analogs, I-palmitoyl-2[(6-pyren-1-yl)]decanoyl-sn-glycero-3-phosphocholine (PPDPC) and 1-palmitoyl-2-[(6-pyren-1-yl)]decanoyl-sn-glycero-3-phosphoglycerol (PPDPG), were obtained from K&V Bioware (Espoo, Finland). Dipalmitoylphosphatidylcholine (DPPC) was from Coatsome (Amagasaki, Hyogo, Japan). The purity of the lipids was checked by thin-layer chromatography on silicic acid-coated plates (Merck), using chloroform/methanol/water (65:25:4, v/v) as a solvent. Examination of the plates after iodine staining or, when appropriate, upon fluorescence illumination revealed no impurities. Lipid stock solutions were made in chloroform. Concentrations of the nonfluorescent phospholipids were determined gravimetrically with a highprecision electrobalance (Cahn Instruments, Cerritos, CA), and those of the pyrene-containing phospholipid analogs were determined spectrophotometrically, using the molar extinction coefficient epsilon = 42,000 at 344 nm. Adriamycin was dissolved in the buffer, and its concentration was determined spectrophotometrically, using epsilon = 11,500 at 480 nm.

Preparation of liposomes

Lipids were mixed in the indicated molar ratios. Unless otherwise indicated, the content of the fluorescent probe (PPDPC or PPDPG) was X = 0.03, which was chosen as a probe concentration to enhance signal-to-noise ratio. In experiments on large unilamellar vesicles (LUVs) with X^sub PG^ >= 0.03, PPDPG (X = 0.03) was preferred to assess more directly the correlation between half-time of the fluorescence decay (t^sub 1/2^) for the binding and the distribution of the probe (representing phosphatidylglycerol, PG) in the membrane. This does not affect the appearance of the fast component, which for fluid membranes was evident only at X^sub PG^ >= 0.04. The lipid mixtures were evaporated to dryness under a stream of nitrogen and then kept under reduced pressure for at least 2 h to remove traces of solvent. Dry residues were hydrated in buffer for 30 min at -15 deg C above the gel -> liquid crystalline transition temperature (T^sub m^) of lipid used. Large unilamellar liposomes (diameter 80 +/- 25 nm; MacDonald et al., 1991) were made by extruding the lipid dispersions 17 times through two Millipore polycarbonate filters (average pore size 100 nm; Millipore, Bedford, MA), using a LiposoFast (Avestin, Ottawa, Canada) small-volume homogenizer. Liposomes were stored in a refrigerator (+4 deg C) and were used within 24 h after hydration.

Stopped-flow fluorescence spectroscopy

Electronic energy is efficiently transferred from a lluorescent energy donor to a suitable acceptor up to a distance of 60 A, with the rate of the transfer depending on the reciprocal of the sixth power of the average distance between the donor and the acceptor. Furthermore, orientation of the donor and acceptor relaxation and excitation dipoles and the magnitude of their spectral overlap determine the efficiency of the energy transfer process (Stryer,1978). Adriamycin has a broad absorption peak centered at 480 nm that overlaps the emission spectrum of pyrene (Mustonen and Kinnunen, 1991). With excitation at 344 nm and measurement of pyrene (donor) emission in the 375-525 nm range, it is possible to measure the binding of adriamycin (acceptor) to lipid membranes embedding pyrene containing a phospholipid analog, e.g., PPDPC or PPDPG (Mustonen and Kinnunen, 1991, 1993; Koiv and Kinnunen, 1994); the membrane association of the drug is readily evident as a decrease in pyrene enussion intensity.

The binding of adriamycin to liposomes was measured using a stoppedflow spectrofluorometer (Otis RSM 1000F; On-Line Instruments, Bogart, GA) equipped with a rapid scanning monochromator and a water-cooled 450 W xenon lamp. Two pneumatically driven (gas pressure 7 bar) syringes were mounted in parallel to inject the reactants into the rapid-mixing quartz glass fluorescence observation chamber (20-mm path length, 1.5-mm diameter) The temperature in the reaction chamber and in the syringes was regulated by a computer-controlled circulating waterbath (Neslab Instruments, Portsmouth, NH). The initial concentration of phospholipid was 50 (mu)M and that of adriamycin 8 (mu)M, thus yielding the corresponding final concentrations of 25 and 4 (mu)M in the mixing chamber. An excitation wavelength of 344 nm was selected with a monochromator, and the beam was passed through subsequent 6.12-mm and 3.16-mm slits. A maximum of 1000 spectra/s was measured by passing the beam through a 1.24-mm slit and a rotating spoke wheel. Fluorescence emission between 375 and 525 nm was recorded with a photomultiplier tube. Photomultiplier output was digitized by a Pentium PC equipped with an AID converter. The fluorescence decays for each sample were monitored by using scanning rates of 1000, 62, and 10 scans/s, with the respecive collection times of 0.25, 5, and 96 s.

Importantly, although the decrease in the fluorescence intensity correlates with the number of adriamycin molecules bound to the liposomes, it is not possible to determine the actual number of drug molecules bound to the vesicles. This is due to the fact that at low surface occupancy, the number of donors quenched by each bound acceptor is large. However, upon increasing surface occupancy by adriamycin, the decrement in emission intensity versus the number of additional drug molecules bound diminishes in a progressive manner. Quantitation of resonance energy transfer processes of this type has been shown to be inherently complex (Drake et al., 1991). Yet, although the above drawback necessarily prevents the construction of an accurate kinetic model and calculation of the true rate constants, it does not compromise in any way the data and the qualitative conclusions of this study.

Steady-state fluorescence measurements

DPPC:PPDPC (97:3, mol/mol) and DPPC:PPDPG (97:3, mol/mol) LUVs for steady-state measurements were prepared as described above. The final lipid concentration was 10 (mu)M. The fluorescence emission spectra were recorded with a Perkin-Elmer LS50B spectrofluorometer with a magnetically stirred cuvette compartment thermostatted with a circulating water bath (WK4DS; Colora Messtechnik Gmbh, Lorch, Germany). The instrument is connected to a Pentium 133-MHz PC, and the data were analyzed with dedicated software from Perkin-Elmer. An excited-state pyrene molecule relaxes to ground state by emitting photons with a maximum at -378 nm (I^sub m^) and the exact peak intensity and spectral fine structure depending on the solvent polarity. During its lifetime, excited-state pyrene can form a characteristic short-lived complex excimer with a ground-state pyrene. This complex relaxes back to ground state by emitting quanta as a broad and featureless band centered at -480 nm (I^sub e^). In the absence of possible quantum mechanical effects, the excimer-to-monomer ratio (I^sub e^/I^sub m^) depends on the rate of collisions between pyrenes. Consequently, for pyrenecontaining phospholipid analogs, the value of I^sub e^/I^sub m^ reflects the lateral diffusion as well as the local concentration of :he fluorophore in the membrane (for a brief recent review see Kinnunen et al., 1993). The excitation and emission bandpasses were set at 2.5 and 5 nm, respectively. Excitation was at 344 nm, and pyrene monomer and excimer emissions were recorded at -378 nm and -480 nm, respectively. The average heating rate in the temperature scanning measurements was 0.1 deg C/min.

Differential scanning calorimetry

LUVs were prepared as above, with the exception that after extrusion the samples were kept at +4 deg C overnight before the heat capacity scans were recorded at a rate of 0.5 deg C/min. Gel ->liquid-crystalline phase transition temperatures of DPPC, DPPC:PPDPC (97:3, mol/mol), and DPPC:PPDPG (97:3, mol/mol) LUVs were recorded by VP-differential scanning calorimetry (VP-DSC) (Microcal, Northampton, MA). This was done and the data were analyzed using the routines of the software provided with the instrument. The final lipid concentration in the DSC cuvette was 0.2 mM. Deviation of the heat capacity from the baseline was taken as the beginning of the transition. Likewise, the point where the enthalpy returned to the baseline was taken as the end of the transition. The endotherms were deconvoluted assuming the components to be Gaussian. It must be stressed, however, that this choice for distribution profiles is arbitrary and stems from the uncertainties regarding the detailed transition mechanisms. Yet, as long as the latter remain poorly understood, the assumption of Gaussian distribution is commonly used (e.g., McMullen et al., 1993). The significance of deconvoluted DSC data must therefore be viewed with caution and can only provide circumstantial support to results from other kinds of measurements. Importantly, the components derived cannot be attributed to physical processes such as phase separation or microdomains.

RESULTS

Dependence on the content of acidic phospholipid in liposomes

Two different binding sites for adriamycin in liposomes and the requirement for acidic phospholipid for the high-affinity membrane association of this drug are well established (Karczmar and Tritton, 1979; Goormagtigh et al., 1980b; Goormaghtigh and Ruysschaert, 1984; Henry et al., 1985; Nicolay et al., 1988; de Wolf et al., 1991; Mustonen and Kinnunen, 1991). Accordingly, it was of interest to determine if these binding sites and the dependence on the content of acidic phospholipids in membranes could also be verified by kinetic measurements. For this purpose POPC was used as a matrix while the content of POPG in the liposomes was progressively increased. Resonance energy transfer between pyrene emission and adriamycin absorption was utilized to monitor the membrane association of the latter (Mustonen and Kinnunen, 1991). The measurement temperature was 30C, so as to have the liposomes in the liquid-crystalline (fluid) state. For liposomes consisting of the zwitterionic POPC and including the pyrene-derivatized phosphatidylcholine analog PPDPC as the fluorescent donor (X^sub PG^ = 0, i.e., with no net charge), the fluorescence decay was single exponential, with a half-time of -2 s (Fig. 2 A). As the mole fraction of PG was increased to 0.04, a second, fast component with t^sub 1/2^ == 2 ms became evident. The amplitude of the fast process increased progressively as X^sub PG^ was increased to 0.25, after which saturation was reached (Fig. 2 C). Concomitantly with the increase in the amplitude of the fast process, the half-time of the slow process increased (Fig. 2 A). Interestingly, t^sub 1/2^ for the fast component varies in a highly irregular manner as a function of X^sub PG^ (Fig. 2 B). More specifically, increasing X^sub PG^ from 0.05 to 0.065 causes t^sub 1/2^ reach a peak. This peak in t^sub 1/2^ is followed by enhanced binding at X^sub PG^ within 0.07-0.09, whereafter a larger value for t^sub 1/2^ at X^sub PG^ = 0.10 is again evident.

To obtain further insight into the nature of the binding process, the content of PG in LUVs was kept constant at X^sub PG^ = 0.25, and ionic strength was varied. In keeping with an electrostatic interaction, an increase in [NaCl] up to 50 mM diminished the amplitude of the fast process, and at 100 mM NaCl this interaction was abolished completely (Fig. 3 B). The half-time of the fast process was nearly independent of [NaCl] and varied between 2 and 4 ms (Fig. 3 A). Instead, t^sub 1/2^ for the slower component varied between 4 and 10 s, and this decay process became faster as [NaCl] was increased to 200 mM.

Ca2+ ions bind to acidic phospholipids and are known to have a strong influence on their phase behavior and organization (Galla and Sackmann, 1975; Hartmann et al., 1977; Eklund et al., 1988, 1991). Binding of adriamycin to LUVs (X^sub PG^ = 0.25) was subsequently measured at increasing [CaCl^sub 2^]. At [Ca2+] from 0 to 150 (mu)M, two-exponential fluorescence decay was observed. The fast component had a t^sub 1/2^ between 2 and 4 ms, and it was not observed when [Ca2+] exceeded 150 (mu)M, thus lending further support to this process being due to an electrostatic attraction. The amplitude of the fast process as a function of clearly revealed a nonlinear dependence (data not shown). More specifically, when [CaCl^sub 2^] was increased from 0 to 25 (mu)M, RFI decreased from 0.76 to 0.25, whereas a further increase in [CaCl^sub 2^ up to 150 (mu)M had a smaller effect, decreasing RFI from 0.25 to 0.11. Increasing [Ca2+] also caused a minor decrease in tlm2 for the slower component, from 12 to 8 s (data not shown). The possibility that vesicle aggregation would have influenced the measured binding can be excluded, as the threshold concentration for the Ca2+-induced aggregation in the literature has been observed to be ~1 mM (Eklund et al., 1991), which is five times the highest [Ca2+] used in the present experiments.

Effects of lateral heterogeneity accompanying the main transition of DPPC/PPDPG LUVs

Exclusion of the pyrene-containing phospholipid analog, PPDPC, from gel-state matrix consisting of DPPC has been shown to result in the formation of microdomains enriched in the fluorescent probe (Somerharju et al., 1985; Hresko et al., 1986). To explore the feasibility of a similar mechanism controlling the lateral distribution of an acidic phospholipid, we included PPDPG at X = 0.03 in a matrix of DPPC. As shown above, with this content of the acidic phospholipid in a fluid bilayer, only the slow binding process is observed, i.e., the surface density of PG in LUVs is too low to yield the fast, electrostatically driven membrane association of adriamycin. By DSC the DPPC/PPDPG LUVs exhibited a single endotherm with a peak centered at 40.4 deg C, thus verifying that this content of PPDPG did not suffice to abolish the main transition of DPPC (see below). In keeping with the predicted segregation of PPDPG into microdomains, the fast component in adriamycin binding was observed at T

DISCUSSION

Confirming the importance of electrostatic attraction between the cationic adriamycin and acidic phospholipids, the rate of the membrane association of this drug was profoundly increased upon the introduction of the negatively charged lipid into fluid POPC LUVs. More specifically, for POPC vesicles the half-time for fluorescence decay was ~2 s, whereas as XPG was increased to >=0.04-s, a fast decay with t^sub 1/2^ in the range of 2-4 ms became evident. Although exact values cannot be calculated from the resonance energy transfer data, this difference in the values for t^sub 1/2^ readily translates into a ~1000-fold enhancement in the binding rate in the presence of the acidic lipid. Reversal of the fast component in the fluorescene decay by increasing [NaCl] and [CaCl^sub 2^] provides unequivocal evidence for this fast process being mediated by electrostatic attraction. With increasing X^sub PG^, t^sub 1/2^ for the slow hydrophobic binding increased from 2 s to 18 s. This is likely to reflect the decrease in the content of adriamycin in the solution, which occurs as a progressively larger fraction of the drug becomes rapidly associated with liposomes by the fast electrostatic mechanism. Charge neutralization in the membrane by the bound adriamycin also readily diminishes the electrostatic attraction between the surface and the drug in the solution. Experiments using different [Adr] demonstrated t^sub 1/2^ for the slow component to be concentration dependent, i.e., to decrease with increasing [Adr] (data not shown). Adriamycin has been shown to penetrate through the lipid bilayers into the liposomes (Praet et al., 1993). Translocation of adriamycin is, however, very slow, requiring 5-10 min (Praet et al., 1993). Accordingly, it is unlikely that this process would contribute to the fluorescence changes reported here. The strong dependence of the fast electrostaticdriven association of adriamycin on X^sub PG^ allows us to use this process to probe the initial organization of the liposome surface prevailing before the binding of adriamycin has commenced. At this point it is relevant to note that we did not study the effects of adriamycin either on the membrane organization or on the transitional behavior of the lipids. Instead, we assessed the effects of changes in membrane organization on the kinetics of binding of adriamycin, with the main emphasis on the fast electrostatic association of the drug with liposomes. More specifically, we measured the half-times for fluorescence intensity decays due to resonance energy transfer (RET), resulting from the electrostatically driven membrane association of drug, as further determined by the membrane content (X^sub PG^), as well as by the initial distribution of the acidic phospholipid, PG, in the membrane. The latter also appears to be strongly dependent on the phase state of the DPPC matrix. After the binding process is complete and the system has reached a new equilibrium, the distribution of PG should be distinctly different from that in the initial state, assessed here. Likewise, the phase behavior of the liposomes has been shown to be altered by the associated drug (Goldman et al., 1978; Constantinides et al., 1986).

An interesting feature of the dependence of t^sub 1/2^ versus X^sub PG^ is that the fast component is absent when X^sub PG^

Our laboratory provided the first evidence for the formation of superlattices in fluid binary alloys (Somerharju et al., 1985). These findings were subsequently confirmed and extended to other compositions (e.g., Tang and Chong, 1992; Chong et al., 1994; Tang et al., 1995). Notably, these superlattices were mainly concluded from steady-state fluorescence measurements, and the underlying repulsive potential was suggested to be steric elastic strain imposed by the bulky fluorescent moities. In the present case it is the strong coulombic repulsion that provides the main driving force. This also involves factors that complicate the prediction of the exact values of XPG producing the superlattices by the acidic headgroups, as follows. In parallel with an increase in XPG, the affinity of the bilayer surface for protons also increases progressively, thus decreasing the degree of the dissociation of the phosphate moiety of PG (Trauble, 1977), with a concomitant reduction in surface charge density. In addition, intermolecular hydrogen bonding between protonated and deprotonated PG headgroups becomes possible (Boggs, 1987), which will also necessarily influence the ordering in the binary lipid alloy and compete with the interaction of adriamycin with the deprotonated acidic phospholipid. The above is in keeping with the membrane binding of another cationic ligand, cytochrome c, becoming (at first glance paradoxically) slower with increasing X^sub PG^ in the liposomes (Subramanian et al., 1998). Obviously, the above processes will also affect the rate of binding of adriamycin to the liposomes. Accordingly, particularly at higher mole fractions of PG, a more complex dependence of ^sub 1/2^ versus X^sub PG^ would be expected. We did not pursue this problem further at this stage.

The strong dependence of the fast process on X^sub PG^ allowed us to use the kinetics of the membrane association of adriamycin to probe changes in the lateral distribution of PG in the course of the phospholipid main transition. Interestingly, at T 0.04), and therefore having increased negative surface density charge. The lack of the slow hydrophobicity-driven binding is likely to reflect the tight packing of the gel-state lipid acyl chains, preventing the intercalation of adriamycin into the bilayer. At T > 34 deg C the slow hydrophobicity driven membrane association of adriamycin also becomes evident. This is likely to result from the formation of "fluid" domains in the bilayer (Jutila and Kinnunen, 1997). The emergence of "fluid" domains in gel bulk is consistent with the enhanced hydrophobic binding of adriamycin to the electrically neutral DPPC/PPDPC LUVs (containing no acidic phospholipid) upon exceeding T == 33 deg C. However, as the slow binding process is observed for the DPPC/PPDPC LUVs already within the temperature range of 30-33 deg C, it is possible that the lack of the slow component for DPPC/ PPDPG LUVs at temperatures lower than 34 deg C is apparent. More specifically, for the latter liposomes the slow component would not be seen, as a large fraction of the drug is already bound to the clustered acidic phospholipid, thus reducing the concentration of adriamycin remaining in the solution after the fast, electostatically driven membrane association of the drug is complete. Above T = 34 deg C the decrease in RFI as a function of T also becomes augmented (Fig. 4 B). In other words, although n (the number of PG molecules encriched in clusters) diminishes when T approaches 34 deg C, as T exceeds this temperature, another process causing a decrease in n becomes effective. The interpretation of the appearance of the slow fluorescence decay at T = 34 deg C as the formation of "fluid" domains in the gel state bilayer is supported by the DSC measurements, which revealed the onset of the main phase transition endotherm at 34 deg C in the DPPC/PPDPG LUVs. Simultaneously with increasing T in the range from 34 to 38 deg C, the PG-enriched clusters seem to be diminishing either in size or number as the amplitude of the fast fluorescence decay is steeply attenuated. Yet, although the number of PG molecules in microdomains is decreasing, the net negative charge density appears to reach a weak local maximum at ~37 deg C, as revealed by the minimum of ~1 ms in t^sub 1/2^.

The fast fluorescence decay for DPPC:PPDPG (97:3, mol/mol) LUVs is not observed at T > 37 deg C. This temperature coincides with the plateau in I^sub e^/I^sub m^ observed in steadystate measurements (Fig. 5), as well as with the onset of peak P in the deconvoluted DSC data for these LUVs (Fig. 6 C and Table 1). The apparent lack of clusters of PPDPG is revealed by the absence of the fast electrostatic component in adriamycin binding above 37 deg C. This is intriguing, as steady-state measurements still reveal highly efficient excimer formation in the temperature range of 38-41 deg C (Fig. 5). We have previously suggested PPDPC to be preferentially accommodated within the interface between gel and "fluid" domains of 1,2-myristoyl-sn-glycero-3-phosphocholine (DMPC) (Jutila and Kinnunen, 1997). Similar behavior for PPDPG in the melting of DPPC matrix would be compatible with the present data. Accordingly, the characteristics of the fast component at 37 deg C (i.e., minimum in both t^sub 1/2^ and amplitude) would result from a change in the dimensionality of the lateral arrangement of PPDPG from 2-D (clusters) to pseudo-1-D (arrays), the latter being formed in the domain boundaries. For the latter, efficient excimer formation is still evident. Importantly, small amounts of adriamycin associated with the PPDPGs concentrated in these 1-D arrays residing within the domain boundaries will strongly diminish the effective charge density and yield a small amplitude for the process, even if the initial affinity is high. An alternative explanation is that the size of the PPDPG clusters falls progressively with T, reaching at ~37 deg C a size no longer having a sufficient number of charges per unit surface area to provide the site for fast binding of adriamycin. However, the latter mechanism does not explain the fact that t^sub 1/2^ has a minimum at 37 deg C. Furthermore, the area of "fluid" domains at 38 deg C, for instance, should be 75% of the total liposome surface area to reduce the concentration of the probe X^sub PG^ below 0.04 and, accordingly, to abolish the fast binding. This seems unlikely because the area under the nondeconvoluted endotherm up to 38 deg C corresponds to only ~ 10-15% of the total enthalpy of the transition. Accordingly, we do consider the former mechanism to be more plausible. The driving force for the enrichment of PPDPG into the boundaries between "fluid" and gel domains in the temperature range of 34-38 deg C can be rationalized as follows (Jutila and Kinnunen, 1997). In brief, a perturbating substitutional impurity of PPDPG is expelled from the gel-state DPPC matrix (Somerharju et al., 1985; Hresko et al., 1986). Upon the transition from gel into the liquid crystalline state, the thickness of the DPPC bilayers is reduced by ~4.5 Angstroms, from 23 to 18.5 Angstroms(Chapman et al., 1967; Lewis and Engelman, 1983). We have previously shown hydrophobic mismatch to promote the clustering of PPDPC in liquid crystalline bilayers of thickness

CONCLUDING REMARKS

The present stopped-flow fluorescence spectroscopy measurements demonstrate that the hydrophobic and electrostatic adriamycin-binding sites in liposomes can be distinguished by their different fluorescence decay half-times. Subsequently, this finding was exploited to study the effect of the DPPC main phase transition on the lateral distribution of a trace amount of an acidic phospholipid probe PPDPG in the bilayer. To our knowledge, this is the first observation that the binding of a drug to a membrane can be controlled by the formation of domains enriched in acidic phospholipids and with high local negative surface charge density. These measurements also provide evidence for the formation of "fluid" domains in the gel state matrix to begin at a temperature well below the peak of the main transition endotherm determined by DSC. Comparison of the stoppedflow data with values of I^sub e^/I^sub m^ derived from steady-state fluorescence measurements provide further evidence for the enrichment of the fluorescent probe, PPDPG, in the interface separating gel-state and fluid domains in the coexistence region present in the course of the transition. These data lend further support to our previous results indicating the length of the interfacial boundary to have a maximum well below the main endotherm, in this case ~3.5 deg below the DSC peak at 40.4 deg C. Accordingly, the models and mechanisms for the phospholipid main transition may have to be revised along the lines discussed previously (Jutila and Kinnunen, 1997). More specifically, our fluorescence data indicate a lack of "fluid"/gel phase boundaries in the range from 38 deg C to 43.7 deg C. Yet, as the peak in heat capacity is observed at 40.4 deg C, the molecular mechanisms underlying the enthalpy changes in this temperature range and the nature of the implied "intermediate" phase pose an intriguing problem. Regarding the latter, it is possible that in this intermediate phase the chain order and lattice translational and positional order are uncoupled, like the liquid-ordered phase present in the phospholipid-cholesterol phase diagram (Vist, 1984; Ipsen et al., 1987; Vist and Davis, 1990).

The technical assistance of Birgitta Rantala is appreciated.

This study was supported by Biocentrum Helsinki and the Finnish State Medical Council.

REFERENCES

Anand, S., H. Verma, L. Kumar, and N. Singh. 1'995. Induction of apoptosis in chronic myelogenous leukemia lymphocytes by hydroxyurea and adriamycin. Cancer Lett. 88:101-105.

Boggs, J. M. 1987. Lipid intermolecular hydrogen bonding: influence on structural organization and membrane function. Biochim. Biophys. Acta. 906:353-404.

Bryant, G., and J. Wolfe. 1989. Can hydration forces induce lateral phase separations in lamellar phases? Eur. Biophys. J. 16:369-374. Chapman, D., R. M. Williams, and B. D. Lad,rooke. 1967. Physical studies of phospholipids. VI. Thermotropic and lyotropic mesomorphism of some 1,2-diacyl-phosphatidylcholines (lecithins). Chem. Phys. Lipids. 1:445-475.

Chong, P. L. G., D. Tang, and I. P. Sugar. 1994. Exploration of physical principles inderlying lipid regular distribution: effects of pressure, temperature, and radius of curvature on ElM dips in pyrene-labeled PC/ DMPC binary mixtures. Biophys. J. 66:2029-2038. Constantinides, P. P., N. Inouchi, T. R. Tritton, A. C. Sartorelli, and J. M. Sturtevant. 1986. A scanning calorimetric study of the interaction of anthracyclines with neutral and acidic phospholipids alone and in binary mixtures. J. BioL Chem. 261:10196-10203.

de Wolf, F. A., M. Maliepaard, F. van Dorsten, I. Berghuis, K. Nicolay, and B. de Kruijff. 1991. Comparable interaction of doxorubicin with various acidic phospholipids results in changes of lipid order and dynamics. Biochim. Biophys. Acta. 1096:67-80. Doroshow, J. H. 1991. Doxorubicin-induced cardiac toxicity. N. Engl. J. Med. 324:843-845.

Drake, J. M., J. Klafter, and P. Levitz. 1991. Chemical and biological microstructures as probed by dynamic processes. Science. 251: 1574-1579.

Duarte-Karim, M., J. M. Ruysschaert, and J. Hildebrand. 1976. Affinity of adriamycin to phospholipids a possible explanation for cardiac mitochondrial lesions. Biochem. Biophys. Res. Commun. 71:658-663. Dumas, F., M. M. Sperotto, M. C. Lebrun, J. 1. Tocanne, and O. G. Mouritsen. 1997. Molecular sorting of lipids by bacteriorhodopsin in dilauroylphosphatidylcholine/distearoylphosphaoidylcholine lipid bilayers. Biophys. J. 73:1940-1953.

Dupou-Cezanne, L., A. M. Sautereau, and J. F. Tocanne. 1989. Localization of adriamycin in model and natural membranes. Eur. J. Biochem. 181:695-702.

Eklund, K. K., and P. K. J. Kinnunen. 1986. Effects of polyamines on the thermotropic behaviour of dipalmitoylphosphatidylglycerol. Chem. Phys. Lipids. 39:109-117.

Eklund, K. K., J. J. Takkunen, and P. K. J. Kinnunen. 1991. Cation-induced aggregation of acidic phospholipid vesicles: the role of fatty acid unsaturation and cholesterol. Chem. Phys. Lipids. 5 :59-66. Eklund, K. K., R. Vuorinen, J. Mikkola, J. A. Virtanen, and P. K. J. Kinnunen. 1988. Ca2+-induced lateral phase separation in phosphatidic acid/phosphatidylcholine monolayers as revealed by fluorescence microscopy. Biochemistry. 27:3433-3437.

Fiallo, M. M. L., and A. Garnier-Suillerot. 1986. Interaction of adriamycin with cardiolipin-containing vesicles. Evidence of an embedded site for the dihydroanthraquinone moiety. Biochim Biophys. Acta. 854: 143-146.

Galla, H.-J., and E. Sackmann. 1975. Chemically induced lipid phase separation in model membranes containing charged lipids: a spin label study. Biochim. Biophys. Acta. 401:509-529.

Gascard, P., M. Sauvage, J-C. Sulpice, and F. Giraud. 1993. Characterization of structural and functional phosphoinositide domains in human erythrocyte membranes. Biochemistry. 32:5941-5948. Goldman, R., T. Facchinetti, D. Bach, A. Raz, and M. Shinitzky. 1978. A differential interaction daunomycin, adriamycin and their derivatives with human erythrocytes and phospholipid bil-yers. Biochim. Biophys. Acta. 512:254-269.

Goormaghtigh, E., R. Brasseur, and J-M. Ruysschaert. 1982. Adriamycin inactivates cytochrome c oxidase by exclusion of the enzyme from its cardiolipin essential environment. Biochem. Biophys. Res. Commun. 104:314-320.

Goormaghtigh, E., P. Chatelain, J. Caspers, and J. M. Ruysschaert. 1980a. Evidence of a specific complex between adriamycin and negativelycharged phospholipids. Biochim. Biophys. Acta. 597:1-14. Goormaghtigh, E., P. Chatelain, J. Caspers, and J. M. Ruysschaert. 1980b. Evidence of a complex between adriamycin derivatives and cardiolipin: possible role in cardiotoxicity. Biochem. Pharmacol. 29:3003-3010.

Goormaghtigh, E., P. Huart, R. Brasseur, and J. M. Ruysschaert. 1986. Mechanism of inhibition of mitochondrial enzymatic complex I-III by adriamycin derivatives. Biochim. Biophys. Acta. 861:83-94. Goormaghtigh, E., and J. M. Ruysschaert. 1984. Anthracycline glycosidemembrane interactions. Biochim. Biophys. Acta. 779:271-288. Hartmann, W., H-J. Galla, and E. Sackmann. 1977. Direct evidence of charge-induced lipid domain structure in model membranes. FEBS Lett. 78:169-172.

Henry, N., E. O. Fantine, J. Bolard, and A. Garnier-Suillerot. 1985. Interaction of adriamycin with negatively charged model membranes: evidence of two types of binding sites. Biochemistry. 24:7085-7092. Holopainen, J. M., J. Y. A. Lehtonen, and P. K. J. Kinnunen. 1997. Lipid microdomains in dimyristoylphosphatidylcholine-ceramide liposomes. Chem. Phys. Lipids. 88:1-13.

Hresko, R. C., I. P. Sugar, Y. Barenholz, and T. E. Thompson. 1986. Lateral distribution of pyrene-labeled phosphatidylcholine in phosphatidylcholine bilayers: fluorescence phase and modulation study. Biochemistry. 25:3813-3823.

Huart, P., R. Brasseur, E. Goormaghtigh, and J-M. Ruysschaert. 1984. Antimitotics induce cardiolipin cluster formation. Possible role in mitochondrial enzyme inactivation. Biochim. Biophys. Acta. 799:199-202. Ipsen, J. H., G. Karlstrom, O. G. Mouritsen, H. Wennerstrom, and M. J. Zuckermann. 1987. Phase equilibria in the phosphatidylcholinecholesterol system. Biochim. Biophys. Acta. 905:162-172. Janmey, P. A., and C. Chaponnier. 1995. Medical aspects of the actin

cytoskeleton. Curr. Opin. Cell Biol. 7:111-117. Jutila, A., and P. K. J. Kinnunen. 1997. Novel features of the main transition of dimyristoylphosphocholine bilayers revealed by fluorescence spectroscopy. J. Phys. Chem. B. 101:7635-7640. Karczmar, G. S., and T. R. Tritton. 1979. The interaction of adriamycin with small unilamellar vesicle liposomes. A fluorescence study. Biochim. Biophys. Acta. 557:306-319.

Kinnunen, P. K. J. 1991. On the principles of functional ordering in biological membranes. Chem. Phys. Lipids. 57:375-399. Kinnunen, P. K. J., A. Koiv, J. Y. A. Lehtonen, and P. Mustonen. 1994. Lipid dynamics and peripheral interactions of proteins with membrane surfaces. Chem. Phys. Lipids. 73:181-207.

Kinnunen, P. K. J., A. Ko iv, and P. Mustonen. 1993. Pyrene-labelled lipids as fluorescent probes in studies on biomembranes and membrane models. In Methods and Applications of Fluorescence Spectroscopy. O. Wolfbeis, editor. Springer Verlag, New York. 159-171. Kinnunen, P. K. J., and P. Laggner. 1991. Phospholipid phase transitions.

Chem. Phys. Lipids. 57:109-399.

Kinnunen, P. K. J., and O. G. Mouritsen. 1994. Functional dynamics of lipids in biomembranes. Chem. Phys. Lipids. 73:1-230. Knoll, W., G. Schmidt, H. Rotzer, T. Henkel, W. Pfeiffer, E. Sackmann, S. Mittler-Neher, and J. Spinke. 1991. Lateral order in binary lipid alloys and its coupling to membrane functions. Chem. Phys. Lipids. 57: 363-374.

Koiv, A., P. Mustonen, and P. K. J. Kinnunen. 1994. Differential scanning calorimetry study on the binding of nucleic acids to dimyristoylphosphatidylcholine-sphingosine liposomes. Chem. Phys. Lipids. 70:1-10. Lee, K. Y. C., J. F. Klingler, and H. M. McConnell. 1994. Electric field-induced concentration gradients in lipid monolayers. Science. 263: 655-658.

Lee, K. Y. C., and H. M. McConnell. 1995. Effect of electric field gradients on the lipid monolayer membranes. Biophys. J. 68:1740-1751. Lehtonen, J. Y. A., J. M. Holopainen, and P. K. J. Kinnunen. 1996. Evidence for the formation of microdomains in liquid crystalline large unilamellar vesicles caused by hydrophobic mismatch of the constituent phospholipids. Biophys. J. 70:1753-1760.

Lehtonen, J. Y. A., and P. K. J. Kinnunen. 1995. Poly(ethylene glycol)induced and temperature-dependent phase separation in fluid binary phospholipid membranes. Biophys. J. 68:525-535.

Lehtonen, J. Y. A., and P. K. J. Kinnunen. 1997. Evidence for phospholipid microdomain formation in liquid crystalline liposomes reconstituted with Escherichia coli lactose permease. Biophys. J. 72:1247-1257. Lewis, B. A., and D. M. Engelman. 1983. Lipid bilayer thickness varies linearly with acyl chain length in fluid phosphatidylcholine vesicles. J. Mol. Biol. 166:211-217.

Ling, Y.-H., W. Priebe, and R. Perez-Soler. 1993. Apoptosis induced by anthracycline antibiotics in P388 parent and multidrug-resistant cells. Cancer Res. 53:1845-1852.

Luan, P., and M. Glaser. 1994. Formation of membrane domains by the envelope proteins of vesicular stomatitis virus. Biochemistry. 33: 4483-4489.

MacDonald, R. C., R. I. MacDonald, B. Ph. P Menco, K. Takeshita, N. K. Subbarao, and L. Hu. 1991. Small volume extrusion apparatus for preparation of large, unilamellar vesicles. Biochim. Biophys. Acta. 1061: 297-303.

Marcie, E. J., and B. R. Lentz. 1986. Phospholipid lateral organization in synthetic membranes as monitored by pyrene- labeled phospholipids: effects of temperature and prothrombin fragment 1 binding. Biochemistry. 25:567-574.

McMullen, T. P. W., R. N. A. H. Lewis, and R. N. McElhaney. 1993. Differential scanning calorimetric study of the effect of cholesterol on the thermotropic phase behavior of a homologous series of linear saturated phosphatidylcholines. Biochemistry. 32:516-522. Momsen, W. E., M. M. Momsen, and H. L. Brockman. 1995. Lipid structural reorganization induced by pancreatic lipase cofactor, procolipase. Biochemistry. 34:7271-7281.

Mou, J., J. Yang, C. Huang, and Z. Shao. 1994. Alcohol induces interdigitated domains in unilamellar phosphatidylcholine bilayers. Biochemistry. 33:9981-9985.

Mouritsen, O. G. 1991. Theoretical models of phospholipid phase transi

tions. Chem. Phys. Lipids. 57:179-194.

Mouritsen, O. G., and M. Bloom. 1984. Mattress model of lipid-protein interactions in mebranes. Biophys. J. 46:141-153. Mouritsen, O. G., and P. K. J. Kinnunen. 1996. Role of lipid organization and dynamics for membrane functionality. In Biological Membranes. K. Merz, Jr., and B. Roux, editors. Birkhauser, Boston. 463-502. Mustonen, P., and P. K. J. Kinnunen. 1991. Activation of phospholipase A2

by adriamycin in vitro. J. Biol. Chem. 266:6302-6307. Mustonen, P., and P. K. J. Kinnunen. 1993. On the reversal by deoxyribonucleic acid of the binding of adriamycin to cardiolipin-containing liposomes. J. Biol. Chem. 268:1074-1080.

Mustonen, P., J. Y. A. Lehtonen, A. Koiv, and P. K. J. Kinnunen. 1993. Effects of sphingosine on peripheral membrane interactions: comparison of adriamycin, cytochrome c, and phospholipase. A2. Biochemistry 26: 2991-2997.

Mustonen, P., J. A. Virtanen, P. J. Somerharju, and P. K. J. Kinnunen. 1987. Binding of cytochrome c to liposomes as revealed by the quenching of fluorescence from pyrene-labeled phospholipids. Biochemistry. 26:2991-2997.

Miller, I., A. Jenner, G. Bruchelt, D. Nietahmmer, .md B. Halliwell. 1997. Effect of concentration on the cytotoxic mechanism of doxorubicinapoptosis and oxidative damage. Biochem. Biophys. Res. Commun. 230:254-257.

Neubig, R. R. 1994. Membrane organization in (3-protein mechanisms. FASEB J. 8:939-946.

Nicolay, K., and B. de Kruijff. 1987. Effects of adriamycin on respiratory chain activities in mitochondria from rat liver, rat heart and bovine heart. Evidence for preferential inhibition of complex III and IV. Biochim. Biophys. Acta. 892:320-330.

Nicolay, K., A-M. Sautereau, J-F. Tocanne, R. Brasseur, P. Huart, J-M. Ruysschaert, and B. de Kruijff. 1988. A comparative model membrane study on structural effects of membrane-active positively charged antitumor drugs. Biochim. Biophys. Acta. 940:197-208. Praet, M., F. Defriese-Quertain, and J-M. Ruysschaert. 1993. Comparison of adriamycin and derivatives uptake into large unilamellar lipid vesicles

in response to a membrane potential. Biochim. Biophys. Acta. 1148: 342-350.

Rodgers, W., and M. Glaser. 1991. Characterization of lipid domains in

erythrocyte membranes. Proc. Natl. Acad. Sci. USA. 88:1364-1368. Rosenoff, S. H., H. M. Olson, D. M. Young, F. Bostick, and R. C. Young. 1975. Adriamycin-induced cardiac damage in the mouse: a small animal model of cardiotoxicity. J. Natl. Cancer. Inst. 55:191-194. Rowe, E. S. 1987. Induction of lateral phase separation in binary lipid mixtures by alcohol. Biochemistry. 26:46-51.

Shimshick, E. J., and H. M. McConnell. 1973. Lateral phase separation in

phospholipid membranes. Biochemistry 12:2351-2360. Skladanowski, A., and J. Konopa. 1993. Adriamycin and daunomycin induce programmed cell death (apoptosis) in tumour cells. Biochem. Pharmacol. 46:375-382.

Soderlund, T., J. Y. A. Lehtonen, J-J. Sievi, and P. K. J. Kinnunen. 1997. Formation of a macroscopic membrane structure by liposomes containing cationic lipids and exposed to DC electric field. Chem. Phys. Lipids. 85:107-114.

Solem, L. E., L. J. Heller, and K. B. Wallace. 1996. Dose-dependent increase in sensitivity to calcium-induced mitochondrial dysfunction and cardiomyocyte cell injury by doxorubicin. J. Mol. Cell. Cardiol. 28: 1023-1032.

Somerharju, P. J., J. A. Virtanen, K. K. Eklund, P. Vainio, and P. K. J. Kinnunen. 1985. 1-Palmitoyl-2-pyrenedecanoyl glycerophospholipids as membrane probes: evidence for regular distribution in liquid-crystalline phosphatidylcholine bilayers. Biochemistry. 24:2773-2781. Stryer, L. 1978. Fluorescence energy transfer as a spectroscopic ruler.

Annu. Rev. Biochem. 47:819-846.

Subramanian, M., A. Jutila, and P. K. J. Kinnunen. 1998. Binding and dissociation of cytochrome c to and from membranes containing acidic phospholipids. Biochemistry. 37:1394-1402. Tang, D., and P. L. G. Chong. 1992. E/M dips. Evidence for lipids regularly distributed into hexagonal super-lattices in pyrene-PC/DMPC dinary mixtures at specific concentrations. Biophys. J. 63:903-910. Tang, D., B. W. Van Der Meer, and S. Y. S. Chen. 1995. Evidence for a regular distribution of cholesterol in phospholipid bilayers from diphenylhexatriene fluorescence. Biophys. J. 68:1944-1951. Tilcock, C. P. S., and P. R. Cullis. 1981. The polymorphic phase behavior of mixed phosphatidylserine-phosphatidylethanolamine model systems as detected by 31P-NMR. Effects of divalent cations and pH. Biochim. Biophys. Acta. 641:189-201.

Tocanne, J.-F., L. Cezanne, A. Lopez, B. Piknova, V. Schram, J.-F. Tournier, and M. Welby. 1994. Lipid domains and lipid/protein interactions in biological membranes. Chem. Phys. Lipids. 73:139-158. Trauble, H. 1977. Membrane electrostatics. In Structure of Biological Membranes. S. Abrahamsson and I. Pascher, editors. Plenum, New York. 509-550.

Vanderkooi, J., M. Erecinska, and B. Chance. 1973. Cytochrome c interaction with membranes. I. Use of a fluorescent chromophore in the study of cytochrome c interaction with artificial and mitochondrial membranes. Arch. Biochem. Biophys. 154:219-229. Virtanen, J. A., P. Somerharju, and P. K. J. Kinnunen. 1988. Prediction of patterns for the regular distribution of soluted guest molecules in liquid crystalline phospholipid membranes. J. Mol. Electron. 4:233-236. Vist, M. R. 1984. Partial phase behavior of perdeuterated dipalmitoylphosphatidylcholine-cholesterol model membranes. MSc Thesis, University of Guelph, Guelph, Ontario, Canada. 103.

Vist, M. R., and J. H. Davis. 1990. Phase equilibria of cholesterol/ dipalmitoylphosphatidylcholine mixtures: 2H nuclear magnetic resonance and differential scanning calorimetry. Biochemistry. 29:451-464. Welti, R., and M. Glaser. 1994. Lipid domains in model and biological

membranes. Chem. Phys. Lipids. 73:121-137. Wu, S. H-W., and H. M. McConnell. 1975. Phase separations in phospholipid membranes. Biochemistry. 14:847-854.

Zaleskis, G., E. Berleth, S. Verstovsek, M. J. Ehrke, and E. Mihich. 1994. Doxorubicin-induced DNA degradation in murine thymocytes. Mol. Pharmacol. 46:901-908.

Tim Soderlund, Arimatti Jutila, and Paavo K. J. Kinnunen

Biomembrane Research Group, Department of Medical Chemistry, Institute of Biomedicine, University of Helsinki, FIN-00014 Helsinki, Finland

Received for publication 26 January 1998 and in final form 22 October 1998.

Address reprint requests to Dr. Paavo K. J. Kinnunen, Department of Medical Chemistry, Institute of Biomedicine, P.O. Box 8, FIN-00014, University of Helsinki, Helsinki, Finland. Tel.: 358-9-1918237; Fax: 3589-1918276; E-mail: paavo.kinnunen@helsinki.fi.

Copyright Biophysical Society Feb 1999
Provided by ProQuest Information and Learning Company. All rights Reserved