Biochimica et Biophysica Acta, 1002 (1989) 109-113 Elsevier BBA 53062
Minimal surface tension, squeeze-out and transition temperatures of binary mi×tures of dipatrnitoylphosphafidy|chofine and unsaturated phosphofipids J. Egberts, H. Sloot and A. Mazure Department of Obstetrics, University Medical Center, Leiden (The Netherlands) (Received 15 July 1988] (Revised manuscript received 14 October 1988)
Key words: Dipalmitoyiphosphatidylcholine; Phospholipid; Squeeze-out: Gel-to-liquid crystal t~ansformatiom Surface activity: Fluorescence polarization
Fluoresence polarization (FP) measurements and surface tension (ST) experiments were performed to determine the gel.to.liquid-crystal transition or melting tempera~we of pb,os|~,holipid mixtures. The FP-temperature diagrams showed main transition temperatures of 41°C for dipalmituylphosphatidylcboline (DPPC). The 7:3 and 9:1 binary mixtures of DPPC and pbosphatidylinositul (Pl), pbosphatidylglycerol (PG) and phosphatidylcholine (PC) had main transition temperatures of, respectively, 32-36 °C and 37-39 o C. The minimal surface tension of DPPC monolayers increased rapidly at 40 °C, suggesting that this was the transition temperature for the melting of these monulayers. This value was in close accordance with the main transition temperature of DPPC, observed with the fluorescence polarization measurements. Melting temperatures of monolayers were higher fer almost all mixtures than the temperatures at which the transition started, indicating preferential squeeze out of the unsaturated component and enrichment of the mono|ayer with DPPC. However, neither the 7:3 D P P C / P C nor the D P P C / P G mixtures could withstand high surface pressures at temperatures above 30 o C, whereas monulayers of D P P C / P G (9:1) became fluid at temperatures above 35 o C. Preferential squeeze-out of the unsaturated phospholipid was especia|ly effective in both the 7:3 and 9:1 D P P C / P l mixtures. These monulayers started to melt at 39-40 ° C, which is above their main transition temperatures of, respectively, 32 and 37 ° C, and which approximate the melting temperature of DPPC. Preferential squeeze-out is essential for an artificial lung surfactant. The estimation of this phenomenon by determining the monolayer melting temperatures is therefore useful for distinguishing between mixtures which are effective surfactants at body temperature and those which are less effective.
Pulmonary surfactant is a phospholipid-protein complex which lines the alveoli and stabilizes the air spaces during expiration. Dipalmitoylphosphatidylcholine (DPPC) is its main component and is essential for withstanding high surface pressures, and thus for preventing alveolar collapse . However, DPPC will not spread spontaneously on water below its gel-to-liquid crystal transition or melting temperature (Tm) of 41°C DPPC, dipalmitoylphosphatidylcholine; PC, phosphatidylcholine; Pl, phosphatidylinositoi; PG, phosphatidylglycerol; FP, fluorescence polarization; Omin, surface tension at minimal surface area; T~ (onset), Tm (main), Te (end) of transition; ST, surface tension. Correspondence: J. Egberts, Department of Obstetrics, University Medical Center Leiden, P.O. Box 9600, 2300 RC Leiden, The Netherlands.
. Mixing DPPC with unsaturated lipids results in a decrease of the Tm and the mixture can then spread at temperatures below 41 ° C. The unsaturated phospholipids phosphatidylcholine (PC), phosphatidylglycerol (PG) or phosphatidylinositol (Pl), present in pulmonary surfactant, function probably as liquefiers of DPPC at body temperature. The spreading of DPPC becomes possible, but these phospholipids may also decrease the ability to withstand high surface pressure unless they are squeezed out of the monolayer during compression [2,3]. The squeeze-out of ungaturated lipids gives rise to the enrichment of DPPC ~,~ the monolayer. Its melting temperature will therefor~ become higher than the transition temperature of the original mixture of DPPC with the unsaturated phospolipids. An increase of the melting temFerature will thus be consistent with the amount of squeeze-out occurring in the monolayer.
0005-2760/89/$03.50 © 1989 Elsevier Science Publishers B.V. (Biomedical Division)
110 Fluorescence polarization measurements and surface tension experiments were therefore performed to determine the main transition temperature or melting of, respectively, fully hydrated mixtures of phospholipids in suspension and of phospholipids in repeatedly compressed and expanded monolayers.
Materials and Methods The phospholipids DPPC, phosphatidylcholine (PC) from egg yolk, phosphatidylglycerol (PG) from egg PC and phosphatidylinositol (Pl) from soybean were purchased from Sigma (St. Louis MO, U.S.A.) The purity was approx. 99~. Each phospholipid migrated as a sin e spot on silica plates with chloroform/ methanol/2-propanol/0.25¢~ KCI/tnethylanune (30: 9: 25: 6: 18, v/v) as the mobile phase  and they were used without further purification. The solvents chloroform, methanol, tetrahydrofuran were analytical grade (Merck, Darmstadt F.R.G.) and 1,6.diphenyl-l,3,5hexatriene was obtained from Koch.Light (Colnbrook, U.K). Double-distilled water was used for the fluorescence polarization measurements and the surface tension experiments.
Fluorescence polarization The fluorescence polarization was measured in a T-shaped fluorometer (excitation wavelength 365 nm) equipped with a temperature.controlled cuvette and with the excitation polarizer set at 0 ° C. The two emission polarizers were set parallel (0°C) and perpendicular (90 o ) (emission 460 nm). The temperature in the cuvette was increased step by step and the emitted fluorescence intensity ( i ) was retarded for 15 s after each change in temperature. The fluorescence polarization (FP) value was calculated aemrding to the formula:
Surface tension measurements A water-and-air-thermostated surfactant balance according to Schoedel et al.  was used for measuring changes in surface tension during compression and expansion of a 100 cmz rhomboid area at 20 s cycles. A relative humidity of 95% or more was maintained during the experiments. The rhomboid area was enclosed by a brass band for preventing leakage of molecules from the area of the monolayer to the area outside the brass band barrier. (Leakage was considered to be absent because the collapse rate of compressed DPPC monolayers was less than 2 m N / m per h) The changes in surface tension during compression and expansion c3'cles were measured using filter paper as Wilhelmy plate (contact angle ~- 0 [9,10]) and the surface temperature was measured by a thermocouple. Changes in temperature (approx. 1 C ° / m i n ) and surface tension were re~orded continuously. Phospholipids in 10-20 /tl chloroform or chloroform/methanol (10: 1, v/v) were spread in excess at 'room' temperature (20-24 ° C) on a surface of deioni~ed and double-distilled water (pH = 6.2-6.6) until the equilibrium surface pressure was reached. A temporary drop of the surface tension, which disappeared within 2 rain, was noticed if pure solvents were 'spread' on the water/air interface. Measurements were therefore started 10-15 rain after spreading of the phospholipids. The temperature (TO [ST]) at which the minimal surface tension started to increase was taken as the onset of melting of the monolayer. Statistical analysis All measurements were performed three or four times for each phospholipid (mixture) and the results are presented as means ± standard deviations. Differences between means were tested by Student's t-test. Results
FP = (no -
no + n o)
The pure phosphofipids or the 9: 1 and 7: 3 mixtures of DPPC and the unsaturated phospholipids in chloroform were dried under N 2 and lipids were suspended in double-distilled water by sonification for 2-5 min at 25 W at 1-4°C. Thereafter, diphenylhesatriene, diluted in water from a stock solution of 1 mM diphenylhexatriene in tetrahydtofuran, was added to the phospholipid vesicles (probe to lipid ratio--1:1000). This mixture was then incubated for 30 min at 37 ° C. The melting temperature (Tin) of the phospholipid/ diphenylhe,xatriene mixture was read from the fluorescence-temperature (FP-T) curve at its point of deflection. The temperatures TO and Te, at which the main gel-to-liquid crystal transition respectively starts and ends, was read at the intersection of the tangents (Fig. la) .
Fig. la shows the relationship between temperature and fluorescence polarization of DPPC and the unsaturated phospholipids PC, Pl and PG. The FP value of DPPC decreased sharply within the temperature range of 39-42 ° C. The deflection point (Tm) was found at 40.9 o C. There was no deflection in the curves of Pl and PG, whereas for PC, an indication of deflection was found at approx. 25 ° C. (A similar deflection at 26°C was observed for phosphatidylethanolamine .) The deflection points of binary mixtures of DPPC and the unsaturated phospholipids are shifted to the left (Fig. lb-d). The 7:3 and 9:1 binary mixtures had main transition temperatures of, respectively, 32-36 °C and of 37.5-39.5°C. (Table I). Wider melting ranges occurred when larger percentages of unsaturated phospholipids were used in the mixtures. The TO [FP] and TO [FP] values of DPPC/PC mixtures were found approx.
100 n mol/ml.
, • X 0 [~
@oR=c X OPPCIPI 9:1 O OPPC I PI 7 : 3
, i i I ! ! DPPC DPPCIPGg:I DPPC/PG 7 : 3 PG
0 o.ao --I
i~. 0.20 0.10 I
Fig. 1. The effect of temperature on the fluorescence polarization of DPPC, PC, PI, PG and binary mixtures of DPPC and the unsaturated phospholipids. The main melting temperature was estimated at the point of deflection (Tin) and the onset and end of the gel-to-liquid crystal transition (To and Te) was estimated from the tangents as is shown for DPPC. Each curve is the mean of three or four independent measurements. The standard deviations were small (approx. the size of the symbols) and are not shown for the sake of clarity.
2 - 4 ° C higher than those of the D P P C / P I or D P P C / P G mixtures. Within the temperature range ¢,f 2 0 - 4 0 ° C , the minimal surface tension (Omi. ) of DPPC approached TABLE I
Melting temperatures of phospholipid mixtures The melting temperatures were determined from the increase of minimal surface tension and by fluorescence polarization measurements (means 4- S.D.). * P < 0.05 (To [ST] > TO [FP]). N = 3 or 4. Phospholipid mixture
Onset of Onset of Main End of increase of transition transition transition omin TO [ST] TO [FP] Tm Te
38.8 4- 0.2 40.9 4- 0.2
42.2 4- 0.4
39.84-0.8* 39.24-0.3 *
D P P C / P G ( 9 : I ) 36.54-0.5* D P P C / P G (7: 3) 30.7 4- 0.8 *
34.74-0.3 26.3 4- 0.3
38.34-0.3 32.9 4- 0.2
40.94-0.3 39.9 + 0.4
Fig. 2. Changes in minimal surface tension of monolayers of binary mixtures of DPPC and the unsaturated phospholipids PC, Pi and PG. The temperature-clependent increase of minimal surface tension of the mixtures D P P C / P C (9:1) (U), DPPC/PI (9:1) (A) and (7:3) (4) (the zone between ~,he thin lines) did not differ significantly from the increase as was found for DPPC (the interrupted line). The minimal surface tension of the mixtures D P P C / P G (7:3) (o) and (9: 1) (O) and D P P C / P C (7 : 3) ((:1) increased at lower temperatures. Each curve is the mean of three or four independent measurements.
zero during compression. The Omin value did not change during continuously repeated compression and expansion of the monolayer. The smallest increase in Omin was observed at 40 o C, the onset of melting of the monolayer (TO [ST]). From that temperature onwards, Omin showed an increase of 2.9 m N / m per C °, but the equilibrium surface tension of 22-24 mN/m, measured at the start of the experiment, was still not approached at 45 o C. The changes of Omi, with temperature of the different phospholipids and mixtures are shown in Fig. 2 and their To [ST] values are given in Table I. The monolayers of DPPC/PC (7 : 3) and of DPFC/PG (7 : 3) started to melt at 30-32°C and high minimal surface tension values of 22-24 m N / m were already found at relatively low temperatures (less than 37°C). These values approximated their surface tension at equilibrium. The DPPC/PC (9:1) mixture, with a TO [FP] of 36.6°C, and both the DPPC/PX (7:3) and (9:1) mixtures, with To [FP] values of, respectively, 27.3 and 34.8°C, had monolayer melting points ranging from 39-41°C and their Omm values increased with higher temperatures. Lower melting points of approx. 34-37 o C were found for monolayers of DPPC/PG (9: 1). Within the temperature range studied (20-45 o C), it was not possible to decrease the Omi, value of the unsaturated phospholipids to values below 25-28 m N / m . In this range, the phospholipids are in a fluid
112 state , which is also shown by their low FP values (Fig. la).
Discussion Two types of methods have been used in this stu,~y for determining the main transition or melting temperature of phospholipid mixtures: fluorescence polarization and surface tension measurements. The large temperature-dependent changes in fluorescence polarization correlate very well with the main transition phase of fully hydrated phospholipids, as determined by differential scanning calorimetry [6,7]. The onset (To [FP]) and end temperature (To [FP]) of the gel.to-liquid crystal transition can also be determined by FP measurements. The absence of a deflectio~ point in the FP.temperaturc curve of the unsaturated phospholipids Pl and PG and the minut deflection in the FP.temperature curve of PC indicate that these phospholipids are already in a fluid state at roora temperature. A 'rigid' monolayer is only present,if the surface temperature is below the melting temperature of the monolayer (To [ST]). This monolayer is in a liquid-condensed phase and further compression will result in a sudden decrease of the minimal surface tension to very low values. The minimal surface tension will then remain constant even after many compression-expansion cycles. A sudden increase of the minimal surface tension occurs only if phospholipids are lost from the monolayer due to their melting (TO [ST]) or because of leakage. The melting point of the monolayer depends on the type of phospholipids and on the pressure within the mono- or bilayer. Compressed mono- or bilayers have a higher melting temperature than the gel-to-liquid crystal transition temperature of the phospholipids under normobaric conditions [13,14]. Goerke and Gonzales  found a melting temperature of 45°C instead of 410 C for a monolayer of DPPC, which was held at low surface tensions or high surface pressures. However, if the monolayer is expanded, the pressure will fall and the monolayer melts at its melting temperature under normobaric conditions. From that moment onwards, the next compression of the monolayer will not lead to an increase in surface pressure and the monolayer melting temperature of, for instance DPPC, will therefore not increase to values above 41 ° C. The TO [FP] is equal to To [ST] if the composition of the monolayer has not changed during compression. Pref--rential squeeze-out of an unsaturated component from ,1 mixture containing DPPC will result in the enrichment of DPPC in the monolayer and of the unsaturated component in bilayer form. These bilayers can be present both in the monolayer as well as in the hypophase . The increase of the monolayer TO [ST] in comparison to TO [FP] is therefore most likely the result of squeeze-out, and the monolayer will thus loose
its 'rigidity' at higher temperature than To [FP]. Our results show that a sudden increase of the minimal surface tension occurs at temperatures specific for each phospholipid mixture. The monolayer cannot withstand high surface pressures anymore, and material is lost from the interface. The possibility exists that compressing and expanding the surface area repeatedly, together with changing the temperature could enhance molecular losses from the monolayer. Monolayer experiments at different constant temperatures were therefore performed and the omi n values were found to be independent of the number of cycles (results not shown). Differences between headgroups and fatty acid tails are important for Tm [6,7,16-19]. Both PC and PG are 'e~-derived' phospholipids, and the shift of Tm of the mixtures from Tm of DPPC was smaller for PC than for PG. This suggests that the headgroup 'glycerol' is responsible for the lower Tm of the DPPC/PG mixtures. However, at pH 4-10, the Tm of DPPG is the same as that of DPPC. [19,20]. Both head and tails are different for (egg) PG and (soybean) Pl, but there were only small differences between the FP melting curves of the 7:3 or the 9:1 mixtures of D P P C / P G or DPPC/PI. From the FP results, it was therefore yet not possible to predict which of the mixtures will lower surface tension to very low values at body temperature. The To [ST] value of most monolayers were higher than their To [FP] values, indicating that preferential squeeze-out occurred in these mixtures. The squeeze-out and probably also the loss of DPPC depend on the type of unsaturated phospholipid . Boonman et al.  showed also that squeeze-out increases with increasing percentages of the unsaturated component. They calculated the molecular loss from mixed monolayers, occurring during the liquid-expanded liquid-condensed transition and found that, if PG > 105, not only PG but also DPPC was lost from the monolayer. With PG < 10~, squeeze out was not detected. However, squeeze-out occurs probably also during the liquid-condensed phase, because we found an increase of TO [ST] compared with To [FP] for the 9:1 mixtures. The DPPC/PG (7: 3) and the (9: 1) monolayets had lost their rigidity at relatively low temperatures, and the minimal surface tension increased consequently. These results suggest that squeeze-out was rather ineffective. Our D P P C / P G (7:3) results are in conflict with those of Morley et al. , but they are similar to those of Yu et al. . DPPC/PG (7:3) mixtures were not very effective in lowering surface tension in the alveoli because gas exchange and lung compliance improved only poorly in in vi~o experiments or clinical trials [25-27]. The DPPC/PC (7:3) mixture lost its monolayer rigidity far below 37 o C, which confirms the results of Metcalfe et al. . This mixture had a TO [ST] value that was almost the same as the TO [FP] value, indicating the absence of preferential squeeze-out.
113 Preferential squeeze-out was very effective in monolayers of the 9:1 DPPC/PC m/xture and especially in those of DPPC and PI. These mixtures showed monolayer stability at temperatures approximating the To of DPPC. This suggests that an almost pure DPPC monolayer was formed. Nevertheless, the main melting temperatures of DPPC/PC (9: 1) and of DPPC/PI (9 : 1) are too high for spreading at 37 o C. The DPPC/PI (7:3) mixture, however, will spread at temperatures belo~v 37 o C, whereas it can be compressed to very low surface tension at temperatures above this critical temperature. This study demonstrates that estimating TO during surface tension measurements is useful in distinguishing mixtures which are ineffective surfactants at body temperature from mixtures which are of interest for further physicochemical and physiological studies. Our results underline the concept of 'preferential squeeze-out'. They also show that the molecular loss of the unsaturated component depends on the type of phospholipid. If squeeze-out is important in the lung, then a DPPC/PI (7:3) mixture might be good basis for making a synthetic lung surfactant [11, 28-30]. Acknowledgement We thank Dr. D.O.E. Gebhardt for his help and critical remarks. References 1 2 3 4
Notter, R.H. and Shapiro, D.L. (1987) Clin. Perinat. 14, 433-479. Bangham, A.D. (1987) Lung 165, 17-25. Watkins, J.C. (1968) Biochim. Biophys. Acta 152, 293-306. Touchstone, J.C., Chen, J.C. and Beaver, K.M. (1982) Lipids 15, 61-62. 5 Shinitzky, M. and Barenholz, Y. (1974) J. Biol. Chem. 249, 2652-2657. 6 Lavaiile, F., Grabielle-Madelmont, C., Petit, J. and Alfsen, A. (1985) Biochemistry 24, 6170-6178. 7 Lentz, B.R., Freire, E. and Bilonen, R.L. (1978) Biochemistry 17, 4474-4480.
8 Schoedel, W., Slama, H. and Hansen, E. (1971) Pfliigers Arch. 322, 336-346. 9 Gieles. P. (i987) Methods of Measurement for the Evaluation :f Monolayer Properties; Development and Applications Ph.D. Thesis, Technical University Eindhoven. 10 Bayramli, E., Van de Ven, T.G.M. and Mason, S.G. (1981) Can. J. Chem. 59, 1954-1962. 11 Yu, S-H., Harding, P.G.R. and Possmayer, F. (1984) Biochim. Biophys. Acta 776, 37-47. 12 Hawco, M.W., Davis, P.J. and Keough K.M.W. (1981) J. Appl. Physiol. 51,509-515. 13 Goerke, J. and Gonzales, J. (1981) J. Appl. Physiol. 51, 1108-1114. 14 Barkai, G., Mashiach, S., Goldman, B., Kalina, M. and Shinitsky, M. (1985) Biochim. Biophys. Acta 834, 103-109. 15 Van Liempd, J.P.J.G., Boonman A.A.H., Demel R.A., ";ieles P.M.C. and Gorree, G.C.M. (1987) Biochim. Biopbys. Acta 82" 495-501. 16 Shimshick, E.J. and McConnell, H.M. (1973) Biochemistry 12, 2351-2360. 17 Lentz, B.R., Barenholz, Y. and Thompson, T.E. (1976) Biochemistry 15, 4529-4537. 18 Keough, K.M.W and Tauesch, H.W. (1986) J. Colloid Interface Sci. 109, 364-375. 19 Szoka, F. and Papahadjopoulos, D. (1980) Annu. Rev. Biophys. Bioeng. 9, 467-508. 20 Watts, A., Harlos, K., Mascke, W. and Marsh, D, (1978) Biochim. 8iophys. Acta 510, 63-74. 2] Boonr t, A.A.H., Snik, A., Egberts, J. and Demel, R. (1984) Prog. Resp. Res. 18, 18-23. 22 Boonman, A.A.H., Machiels, F.H.J., Snik, A.F.M. and Egberts, J. (1987) J. Colloid Interface Sci. 120, 456-468. 23 Morley, C. and Bangham, A. (1981) Prog. Resp. Res. 15, 261-268. 24 Yu, S-H., Chung, W., Oiafson, R.W., Harding, P.G.R. and Possmayer, F. (1987) Biochim. Biophys. Acta 921,437-448. 25 Egan, E.A., Notter, R.H., Kwong, M.S. and Shapiro, D.L. (1983) J. Appl. Physiol. 55, 875-883. 26 Ten Centre Study Group (1987) Brit. Med. J. 294, 991-996. 27 Wilkinson, A., Jenkins, P.A. and Jeffrey, J.A. (1985) Lancet ii, 287-291. 28 Metcaife, i.L., Enhorning, G. and Possmayer F. (1980) J. Appl. Physiol. 49, 34-41. 29 Durand, D.J., Clyman, K.l., Heymann, M.A., Clements, J.A., Mauray, F., Kitterman, J. and Ballard, P. (1985) J. Pediatr. 107, 775-780. 30 Turcotte, J.G., Sacco, A.M., Steim, J.M., Tabak, S.A. and Notter, R.H. (1977) Biochim. Biophys. Acta 488, 235-248.