Collapse and Shedding Transitions in Binary Lipid Monolayers

Langmuir 2006, 22, 2993-2999

2993

Collapse and Shedding Transitions in Binary Lipid Monolayers Coating Microbubbles Gang Pu,† Mark A. Borden,‡ and Marjorie L. Longo*,† Department of Chemical Engineering & Materials Science and Department of Biomedical Engineering, UniVersity of California, DaVis, California 95616 ReceiVed NoVember 10, 2005. In Final Form: January 12, 2006 We report on a fluorescence microscopy study of the monolayer collapse and shedding behavior due to shell compression during the dissolution of air-filled, lipid-coated microbubbles in degassed media. The monolayer shell was comprised of saturated diacyl phosphatidylcholine (C12:0 to C22:0) and an emulsifier, poly(ethylene glycol)-40 stearate. The morphologies of monolayer collapse structures and shed particles were monitored as a function of phospholipid acyl chain length (n) and temperature. The two components formed a single miscible phase when the phospholipid was near or above its main phase transition temperature, and collapse occurred via suboptical particles to vesicles (both were shed) and tubes as chain length increased. Conversely, two-phase coexistence was observed when the lipid was below its main phase transition temperature. For these bubbles, a transition from primary collapse to secondary collapse was observed. Primary collapse was observed as a loss of expanded phase due to vesiculation. Secondary collapse involved the rapid propagation of monolayer folds and simultaneous deformation. For very rigid monolayers, we observed substantial surface buckling with simultaneous nucleation and growth of folds. The folds merged at a single point or region, providing a conduit for the entire excess lipid to shed in a single event, and the bubble smoothed and became more spherical. These results are discussed in the context of general binary phospholipid collapse behavior, microbubble dissolution behavior, medical applications, and the dissolution behavior of natural microbubbles.

Introduction Lipid-coated microbubbles are currently being used in contrast echocardiography1 and show great potential for radiological applications, including molecular imaging.2-4 In addition, lipidcoated microbubbles are being developed for therapeutic applications,5 such as ultrasound-assisted targeted drug delivery, thermal ablation with high-intensity focused ultrasound, and metabolic gas transport. In these applications, preformed microbubbles are injected intravenously, and circulation time is on the order of tens of minutes.6,7 The mechanisms underpinning elimination are not well defined. However, the general consensus is that non-ultrasound-induced microbubble destruction occurs through dissolution of the gas core into the blood pool.8 There is interest in gaining knowledge about the fate of the lipid shell during this process, as it will aid in the rational design of techniques to increase microbubble viability and reduce unwanted bioeffects. Of additional interest is the collapse and shedding behavior of the shells of naturally occurring microbubbles. Such bubbles profoundly affect the mechanical, acoustic, and transport properties of seawater.9 The fate of the shell during dissolution, * Corresponding author. E-mail: [email protected]. † Department of Chemical Engineering & Materials Science. ‡ Department of Biomedical Engineering. (1) Tsutsui, J. M.; Kusler, M.; Porter, T. R. Curr. Opin. Cardiol. 2005, 20, 381. (2) Lindner, J. R. Nat. ReV. Drug DiscoVery 2004, 3, 527. (3) Klibanov, A. L. Bioconjugate Chem. 2005, 16, 9. (4) Bloch, S. H.; Dayton, P. A.; Ferrara, K. W. IEEE Eng. Med. Biol. Mag. 2004, 23, 18. (5) Unger, E. C.; Porter, T.; Culp, W.; Labell, R.; Matsunaga, T.; Zutshi, R. AdV. Drug DeliVery ReV. 2004, 56, 1291. (6) Fritz, T. A.; Unger, E. C.; Sutherland, G.; Sahn, D. InVest. Radiol. 1997, 32, 735. (7) Klibanov, A. L. AdV. Drug DeliVery ReV. 1999, 37, 139. (8) Kabalnov, A.; Bradley, J.; Flaim, S.; Klein, D.; Pelura, T.; Peters, B.; Otto, S.; Reynolds, J.; Schutt, E.; Weers, J. Ultrasound Med. Biol. 1998, 24, 751. (9) D’Arrigo, J. S. Stable Gas-in-Liquid Emulsions: Production in Natural Waters and Artificial Media; Elsevier Science Publishing Co.: New York, 1986; Vol. 40.

therefore, is of consequence to many areas of inquiry, including marine biology, sonar technology, and atmospheric science, to name a few.10 Typically, two components are necessary to form the monolayer shell of a synthetic microbubble.11 The primary species is a long-chain lipid, typically saturated diacyl phosphatidylcholine (DinPC; where n equals the number of carbons per acyl chain), that imparts low surface tension,12 high mechanical stability,13 and low gas permeability.14,15 The secondary component is an emulsifier, typically containing poly(ethylene glycol) (PEG). This two-component system is analogous to the composition of naturally occurring microbubbles, which contain acyl lipids and degraded glycoproteins.9 A common monolayer mixture used to form synthetic microbubbles is DinPC and PEG-40 stearate (PEG40S).11,13-18 We previously studied the gas-core dissolution behavior and quiescent shell phase behavior of this particular system for a range of acyl chain lengths (DinPC, n ) 12, 14, 16, 18, 20, 22, and 24).17,18 As n was increased, the bubble morphology during dissolution changed from spherical, to deformed sphere with a stepping down of size, to cycles of surface bucklingsmoothing (the smoothing defined here as the spontaneous surface smoothing and partial or full restoration of microbubble sphericity). Similar stepping down in size and deformation has been observed in dissolving seawater microbubbles as well.10,19 (10) Johnson, B. D.; Cooke, R. C. Science 1981, 213, 209. (11) Klibanov, A. L. Ultrasound contrast agents: Development of the field and current status. In Contrast Agents II; Krause, W., Ed.; Springer-Verlag: New York, 2002; Vol. 222, p 73. (12) Duncan, P. B.; Needham, D. Langmuir 2004, 20, 2567. (13) Kim, D. H.; Costello, M. J.; Duncan, P. B.; Needham, D. Langmuir 2003, 19, 8455. (14) Borden, M. A.; Longo, M. L. J. Phys. Chem. B 2004, 108, 6009. (15) Pu, G.; Longo, M. L.; Borden, M. A. J. Am. Chem. Soc. 2005, 127, 6524. (16) Kim, D. H.; Klibanov, A. L.; Needham, D., Langmuir 2000, 16, 2808. (17) Borden, M. A.; Longo, M. L. Langmuir 2002, 18, 9225. (18) Borden, M. A.; Pu, G.; Runner, G. J.; Longo, M. L. Colloids Surf., B 2004, 35, 209. (19) Johnson, B. D.; Cooke, R. C. Limnol. Oceanogr. 1980, 25, 653.

10.1021/la0530337 CCC: $33.50 © 2006 American Chemical Society Published on Web 02/25/2006

2994 Langmuir, Vol. 22, No. 7, 2006

Direct observation of the shell during these events would provide evidence for the mechanism behind this remarkable phenomenon. Interestingly, quiescent synthetic microbubbles exhibited twophase coexistence. We interpreted the Langmuir pressure-area (Π-A) isotherm and fluorescence microscopy results as lateral phase separation between DinPC (T < Tm; Tm ) fluid-gel transition temperature for the lipid in its bilayer state) and the emulsifier, PEG40S. The question arises as to how such phase separation might affect the collapsing shell. During gas-core dissolution, the monolayer shell of the microbubble undergoes compression and ultimately collapses. In general, as observed on the Langmuir film balance,20-23 reduction in the area available to molecules at the air-water interface causes a series of two-dimensional (2D) phase transformations20-23 (gaslike (G) state f expanded (E) phase f anisotropic condensed (C) phase). Reduction in area below that found at the equilibrium spreading pressure drives the monolayer into a metastable state from which it relaxes by the nucleation and growth of three-dimensional (3D) structures.24-27 This 2Dto-3D relaxation phenomenon is referred to as collapse, and the surface pressure (Π) at which this occurs on the laboratory time scale is termed the collapse pressure (Πc). Collapse results in 3D defects and discontinuities in the 2D monolayer and therefore determines the resultant morphology of the overcompressed film. A variety of 3D structures have been observed during monolayer collapse. These include disklike structures that grow and detach,28 nanometer-scale vesicles and folds,29,30 micronscale vesicles,31,32 buckling,33,34 and micron-scale folds and bundles.30,32,35 In an attempt to provide an experimental and theoretical framework to explain often-disparate experimental observations, Lee and co-workers32,36 studied a binary phospholipid monolayer consisting of 70 mol % saturated 1,2-diacylsn-glycero-3-phosphatidylcholine (n ) 16 carbons per acyl chain, Di16PC) and 30 mol % singly unsaturated 1-acyl,2-acyl-snglycero-3-phosphatidylglycerol (n ) 16:0-18:1, 16:0-18: 1PG).32,36 They varied temperature to alter phase equilibrium and found a strong correlation between the temperature-dependent morphology of the monolayer and its 2D-3D structure. It was observed that, at approximately the critical temperature for phase separation of the two components (33 °C) in which both were in the E phase, collapse occurred through 1-µm vesicle budding and detachment/shedding and tubes that remained attached. A few degrees above the critical temperature, collapse occurred only through the budding and shedding of 1-µm vesicles. Below this critical temperature, collapse occurred through vesiculation (20) Mohwald, H. Annu. ReV. Phys. Chem. 1990, 41, 441. (21) McConnell, H. M. Annu. ReV. Phys. Chem. 1991, 42, 171. (22) Weis, R. M. Chem. Phys. Lipids 1991, 57, 227. (23) Knobler, C. M.; Desai, R. C. Annu. ReV. Phys. Chem. 1992, 43, 207. (24) Gaines, G. L., Jr. Insoluble Monolayers at Liquid-Gas Interfaces; Interscience Publishers: New York, 1966. (25) Smith, R. D.; Berg, J. C. J. Colloid Interface Sci. 1980, 74, 273. (26) Nikomarov, E. S., Langmuir 1990, 6, 410. (27) Vollhardt, D. AdV. Colloid Interface Sci. 1993, 47, 1. (28) Schief, W. R.; Touryan, L.; Hall, S. B.; Vogel, V. J. Phys. Chem. B 2000, 104, 7388. (29) Ridsdale, R. A.; Palaniyar, N.; Possmayer, F.; Harauz, G. J. Membr. Biol. 2001, 180, 21. (30) Saccani, J.; Castano, S.; Beaurain, F.; Laguerre, M.; Desbat, B., Langmuir 2004, 20, 9190. (31) Alig, T. F.; Warriner, H. E.; Lee, L.; Zasadzinski, J. A. Biophys. J. 2004, 86, 897. (32) Gopal, A.; Lee, K. Y. C. J. Phys. Chem. B 2001, 105, 10348. (33) Saint-Jalmes, A.; Gallet, F. Euro. Phys. J. B 1998, 2, 489. (34) Saintjalmes, A.; Graner, F.; Gallet, F.; Houchmandzadeh, B. Europhys. Lett. 1994, 28, 565. (35) Lipp, M. M.; Lee, K. Y. C.; Takamoto, D. Y.; Zasadzinski, J. A.; Waring, A. J. Phys. ReV. Lett. 1998, 81, 1650. (36) Diamant, H.; Witten, T. A.; Ege, C.; Gopal, A.; Lee, K. Y. C. Phys. ReV. E 2001, 36, 6306.

Pu et al.

(attached) at a lower Π (near 55 mN/m) and folding at higher Π (near 72 mN/m). Folds were observed to rapidly propagate (up to millimeters in length in 30-100 ms) and showed no observable preference for the domain boundaries. Below ∼28 °C, collapse occurred through the formation of reversible folds only at high pressure. These changes were viewed as being driven by a decrease in the miscibility between the two components, which, in turn, led to changes in local composition, morphology, and mechanical properties. In these previous observations, folding or vesiculation in biphasic monolayers originated mostly at the boundaries between the E-phase and C-phase domains. Qualitatively, the authors proposed a model suggesting that, whenever there are two phases in coexistence, the difference in the spontaneous curvature and rigidity of the two phases induces the formation of slightly elevated “mesas”. Their model suggests that the boundaries between the C and E phases are the locations where the monolayer is inflected and are therefore the natural locations for nucleation of collapse. Herein, we investigate with fluorescence microscopy the 2D and 3D behavior of the monolayer shell, comprising DinPC and PEG40S, of dissolving microbubbles. The microbubble shell provides an advantageous system on which to observe these phenomena because it is continuous and becomes fully compressed because of the effects of Laplace overpressure, which is a manifestation of surface tension on a curved interface.37,38 Both phospholipid acyl chain length and temperature were varied to encompass a wide range of surface morphologies. Direct observation of the shell allowed us to discern the effects of shell morphology on collapse and shedding behavior. We note general trends that support the previous work of Lee and co-workers32,36 on two-phase, binary Langmuir monolayers. This was the case, even though the area covered by the monolayers on the microbubbles was several orders of magnitude less than the monolayers typically used in Langmuir troughs. Important differences also existed as a result of the different (and wider range) of components used here. For example, the use of a singlechained emulsifier (PEG40S) versus the 16:0-18:1PG used in the studies of Lee and co-workers resulted in a difference in the collapse behavior at temperatures significantly below the miscibility temperature of the two components based on Tm. We also provide direct evidence for the mechanism underlying the stepping down in size and buckling-smoothing effect previously observed in microbubbles formed by the same method.17 The implications of our results are discussed in the context of medical and natural microbubbles. Materials and Methods Materials. The lipid component, saturated 1,2-diacyl-sn-glycerol3-phosphatidylcholine (n)12, 14, 16, 18, 20, 22 carbons per acyl chain; DinPC), and fluorescent probe, 1-palmitoyl-2-[12-[(7-nitro2-1,3-benzoxadiazol-4-yl)amino]dodecanoyl]-sn-glycero-3- phosphocholine (NBD-PC), were purchased from Avanti Polar Lipids, Inc. (Alabaster, AL). The fluorescent probe, 3H-indolium,2-[3-(1,3dihydro-3,3-dimethyl-1-octadecyl-2H-indol-2- ylidene)-1-propenyl]3,3-dimethyl-1-octadecyl-perchorate (DiI), was purchased from Molecular Probes (Eugene, OR). The emulsifier, PEG40S, was purchased from Sigma (St. Louis, MO). Deionized water was purified to 18 MΩ cm with a Barnstead Nanopure water system (Dubuque, IA). Microbubble Preparation and Dissolution Experiments. Microbubbles were prepared as described elsewhere.17 Briefly, each mixture for a microbubble solution contained 89 mol % DinPC, 10 (37) de Laplace, P. S. Mecanique Celeste, Supplement to Book 10; Bachelier: Paris, 1806. (38) Young, T. Miscallaneous Works. J. Murry: London, 1855; Vol. I.

Microbubble Shell Collapse BehaVior

Langmuir, Vol. 22, No. 7, 2006 2995 Table 1. Summary of Microscopy Observations.a

b

macroscopic dissolution behavior primary secondary

n

Tm (°C)

T (°C)

12 14

-1 23

22 5 22

uniform large domains uniform

smooth smooth smooth

deformed

16

41

22

smooth

deformed

18

55

52 22

large, medium, small domains uniform medium, small domains

20

66

22

small domains

smooth

22

75

22

small domains

smooth

microstructurec

smooth smooth

Result and Discussion Air-filled microbubbles coated with an emulsifier (PEG40S) and a saturated phospholipid of various chain lengths (DinPC) were formed through high-power tip sonification of a vesicle/ micelle aqueous solution. This process has been documented to heat the solution up to approximately 60 °C.18 The contents then were cooled to RT () 22 ( 1 °C) on the bench under ambient conditions. A fluorescent probe was included that partitioned

suboptical small vesicles small vesicles + tubes small vesicles large vesicles small vesicles

deformed or deformed + buckling-smooth deformed + buckling-smooth deformed + buckling-smooth

a Microbubble shell monolayers consist of DinPC + PEG40S. b Values taken from J. R. Silvius.39 < 2 µm2. d Small < 1 µm; large > 2 µm.

mol % PEG40S, and 1 mol % NBD-PC. The mixture was dissolved in chloroform and mixed in a 25 mL glass scintillation vial. The solution was dried under nitrogen gas and put in a vacuum oven for 2 h at a temperature at least 5 °C above the lipid phase transition temperature. Then, the thin lipid-emulsifier film was rehydrated with Nanopure water via a bath sonicator (Fisher, Pittsburgh, PA) with heat for 20 min. The final concentration for the lipid solution was 3 mg/mL. The microbubbles were made immediately after bath sonication. The solution was first dispersed with a tip sonicator at low power (Branson, Danbury, CT) for 3 min. Then the tip sonicator was placed at the air-water interface to entrain air and form microbubbles at high power for 10 s. Alternatively, a brighter probe, DiI, was used with identical results. In this case, DiI was added directly to microbubbles made through sonication of a suspension of small vesicles (90 mol % DinPC and 10 mol % PEG40S). The volume of DiI added was 3 µL of a 1 M solution in pure ethanol for every 1 mL of initial vesicle suspension. After microbubble formation, the suspension was cooled to room temperature (RT) in the vial under ambient conditions. The dissolution experiments were performed similarly to previous experiments.17 Nanopure water was degassed while undergoing vigorous stirring. Measurements by a dissolved oxygen meter (model YSI 556 MPS, Multi Probe System; YSI Incorporated, Yellow Springs, OH) revealed that the partial pressure of the oxygen in the surrounding medium to that at saturation was approximately 1/2. Microbubbles were injected from a sample syringe into a laminar flow perfusion chamber (Warner Instruments, Hamden, CT). Several chamber volumes of degassed water were then pumped through the chamber via a syringe pump (Kd Scientific, Holliston, MA) to purge the system. For the experiments performed at 5 and 52 °C, the perfusion chamber was placed on a heating plate, and the temperature was controlled by a heater/chiller circulation bath (Thermo-Neslab, Portsmouth, NH). The degassed water was also heated or cooled to the same temperature before being pumped into the perfusion chamber. Microbubbles (which rise to the top of the chamber) were imaged from the top or bottom of the chamber as they dissolved into degassed water using a Nikon eclipse e400 or Diaphot 300 microscope (Nikon, Melville, NY), respectively, equipped with a high-resolution Orca digital camera (Hamamatsu, Japan).

microscopic shell collapse and shedding behavior primaryd secondary

small vesicles small vesicles c

folds or vesicles folds or vesicles folds or bucklefolds-smooth bucklefolds-smooth bucklefolds-smooth

Large > 25 µm2; medium ) 2-25 µm2; small

Figure 1. Fluorescent videomicrographs of E-phase microbubble shell collapse and shedding behaviors during dissolution in degassed media. (A) Di12PC + PEG40S microbubble at RT. (B) Di14PC + PEG40S microbubble at RT. (C) Di14PC + PEG40S microbubble at RT, alternative mechanism. (D) Di16PC + PEG40S microbubble at 52 °C. Scale bars represent 20 µm.

into the E phase; C phases appeared dark. A significant population of bubbles coated with DinPC below its Tm exhibited both C and E phases in the form of dark domains (C phase) surrounded by a bright matrix (E phase). Previous fluorescence microscopy and Langmuir Π-A isotherm results of monolayers indicated that dark domains were primarily composed of DinPC, and the bright fluorescent regions were enriched in PEG40S.18 Dissolution of the air core was induced by introduction of degassed water. Dissolution resulted in a decrease in the area of the air-water interface of each bubble, and, thus, we observed the impact of induced compression on the 2D and 3D surface morphology. The results, as summarized in Table 1, are discussed in reference to previous results by Lee and co-workers32,36 for binary Langmuir monolayers of 16:0-18:1PG + Di16PC and our previous work17,18 on surface morphology and dissolution of microbubbles coated with the same components used here. Collapse and Shedding of Microbubble Shells Containing Two E-Phase Components. Figure 1 shows typical dissolution behaviors of microbubbles coated with DinPC + PEG40S where T J Tm of the DinPC component (n ) 12, 14 at RT, and n )

2996 Langmuir, Vol. 22, No. 7, 2006

16 at 52 °C). The surfaces of all of these bubbles were uniformly fluorescent during the bubble dissolution, indicating that the monolayers were in the E phase. Uniform fluorescence was also observed for all aggregates formed. As observed previously in bright field, these microbubbles remained spherical as they dissolved into degassed water.17 Figure 1A shows the RT dissolution behavior seen in all Di12PC + PEG40S microbubbles. No fluorescent material was visualized departing from the bubble surface. This suggests that micelle-sized aggregates were shed from the surface. This is not surprising since Di12PC is quite hydrophilic (shortening by 2 carbons eliminates vesicle formation) and PEG40S forms micelles. Figure 1B and Movie 1 (Supporting Information) show the RT dissolution typical of Di14PC + PEG40S microbubbles. Small fluorescent aggregates, presumably vesicles, with a fluorescent diameter less than 1 µm, were continuously shed from the surface. For these collapse and shedding processes, the rate of aggregate formation, growth, and detachment accelerated as the size of the microbubble decreased. Some fluorescent shell remnants were found after the gas core had shrunk to 1-µm scale. As shown in Figure 1C and Movie 2 (Supporting Information), an alternative collapse mechanism was observed (∼35% of ∼800 microbubbles observed) for Di14PC + PEG40S microbubbles. During dissolution, flexible tubular structures appeared on the shell surface. These tubes were attached to the shell at one end and were observed to grow as the gas core dissolved. Single tubes could reach a length of 100 µm. This behavior was remarkably similar to the collapse and shedding behavior of the Di16PC + 16:0-18:1PG (7:3) monolayers at 37 °C observed by Gopal and Lee.32 Both budding vesicles that detach and flexible tubular structures were observed. In comparison, the reduced temperatures of the PC lipids were almost identical, that is, just a couple of degrees below Tm. As is pointed out by Lee and co-workers,32,36 the tubular structures were predicted by Hu and Granek.40 A hexagonal array of fingerlike tubes is favored over the flat monolayer for a uniform monolayer with a nonzero spontaneous curvature above a critical Π. It is pointed out that tubular collapse structures are unlikely in single-component monolayers, but that the presence of more than one component might stabilize these structures. By including either 16:0-18:1PG, with its charged headgroup, or PEG40S, with its bulky PEG headgroup, we would expect that the energy associated with the highly curved tubular surfaces would be decreased in comparison to a single-component phosphatidylcholine lipid system. This speculation that PEG40S is contained in the collapse structures agrees with previous Π-A isotherms in which only one Πc (∼50 mN/m) was observed for monolayers of this mixture,18 indicating that PEG40S is not squeezed out at a pressure below the Πc of Di14PC. (The Πc of a monolayer of PEG40S at RT is approximately 35 mN/m.18) In contrast to the work on Di16PC + 16:0-18:1PG at 37 °C in which both shed vesicles and tubes were observed in coexistence, bubbles coated with Di14PC + PEG40S at RT produced either tubes or vesicles. This may be a result of the lack of control of exact composition for the microbubble shells; therefore, some shells have more PEG40S in comparison to others. Similar dissolution behavior also occurred for all Di16PC + PEG40S microbubbles at 52 °C (i.e., above Tm ∼ 41 °C for Di16PC), as shown in Figure 1D. Vesicles, as opposed to tubes, were observed to bud from the monolayer shell. These vesicles were generally circular in shape and could reach up to 5 µm in (39) Silvius, J. R. Lipid-Protein Interactions. John Wiley & Sons: New York, 1982. (40) Hu, J. G.; Granek, R. J. Phys. II 1996, 6, 999.

Pu et al.

Figure 2. Fluorescence videomicrographs of coexisting E- and C-phase microbubble shell collapse and shedding during dissolution in degassed media. (A) Primary collapse from a Di16PC + PEG40S microbubble at RT. Arrow points to typical budding vesicle at the boundary of the E-phase and C-phase domains. (B) Secondary collapse from a Di16PC + PEG40S microbubble at RT. Thin lines indicate the location of an impending fold. Numbers indicate folding domains for Table 2. Arrows point to growing collapsed E phase during fold formation. (C) Di16PC + PEG40S microbubble at RT. (D) Di18PC + PEG40S microbubble at RT. The lines of arrows in frames 2 and 4 are adjacent to a fold that has recently formed in each image. (E) Di20PC + PEG40S microbubble at RT. Arrows point to the region where folds meet. (F) Bright-field image of secondary collapse and shedding from a Di20PC + PEG40S microbubble at RT. Inset shows a bubble with measurable change in roundness resulting from shedding. Scale bars represent 20 µm.

diameter, a much larger size in comparison to the vesicles discussed above. Perhaps, this reflects the relatively larger thermal fluctuations at 52 °C compared to those at RT in enhancing the vesicle swelling. Intermittent vesicle detachment from the monolayer shell was observed. Attached or free vesicles were deformed into tubular shapes under the influence of shear flow induced by thermal gradients, and the tubes were observed to reach hundreds of microns in length. Collapse and Shedding of Microbubble Shells Containing E-Phase and C-Phase Components. We increased the lipid chain length (n ) 16, 18, 20, 22) at RT and decreased temperature (for shells containing Di14PC) such that the temperature at which we observed bubble dissolution was significantly below the Tm of the DinPC component. Again, we evaluate our result with the framework set out by Lee and co-workers32,36 and our previous observations on microbubbles.17,18 Figure 2 shows the typical dissolution behaviors of microbubbles coated with DinPC + PEG40S where T < Tm of the DinPC component. Primary Stage: Vesiculation (Attached). For DinPC (n e 18) coated bubbles, at the beginning of dissolution, a significant

Microbubble Shell Collapse BehaVior

Langmuir, Vol. 22, No. 7, 2006 2997 Table 2. Domain Area Change from Figure 2B, Resulting from Surface Folds

Figure 3. Cartoons of collapse and shedding mechanisms. (A) Collapse through vesiculation, corresponding to Figure 2A. (B) Collapse through folding, corresponding to Figure 2B-D. It is demonstrated that folds run through C domains (black) as well as intervening E regions (grey). (C) Surface buckling (waviness) and simultaneous fold formation (top) followed by shedding of the excess monolayer (bottom), corresponding to Figure 2E,F. The fold serves as a conduit to shed the excess monolayer.

population of bubbles contained visible (1-10 µm) dark domains surrounded by a bright matrix. Previous fluorescence microscopy and Π-A isotherms of Langmuir monolayer results indicated that dark domains were C-phase phospholipids and the bright fluorescent regions were E phase enriched in PEG40S.18 As the bubbles decreased in size, the surface area of the bright matrix decreased, while the overall area of the domains remained unchanged. When the areas of the two regions were approximately the same, bright spots of size less than 1 µm appeared at the borders between the C-phase domains and the surrounding E phase. Figure 2A demonstrates the first stage of dissolution for bubbles coated with Di16PC + PEG40S observed at RT, that is, the loss of the bright matrix area and the appearance of bright spots. These aggregates, presumably unilamellar vesicles, increased in concentration until their fluorescence merged to make the E phase brighter. The nucleation of attached vesicles at the domain borders supports the experimental and theoretical observations by Lee and co-workers. 32,36 In fact, height mesas, which are theoretically proposed to be the location of collapse nucleation, have been observed using freeze-fracture and optical microscopy on the surfaces of microbubbles stabilized by the component used here.13,17 In Figure 3A, we provide a cartoon of the location of attached vesicles. During this primary collapse stage, the bubbles dissolved smoothly, remained circular, and no detachment of material was observed. As chain-length increased, the percentage of bubbles with microscopically visible domains decreased. This was a result of the processing method used to form the bubbles in which they were cooled following bubble formation by sonication. Longer-chain-length lipids have lower energy nuclei at RT, and thus smaller domains were formed. This observation is in basic agreement with the results of our previous study in which bubbles coated with lipids of increasing chain length and PEG40S were treated with a consistent heating and cooling schedule.15 Bubbles with subvisible domains tended to appear mottled. In this case, gas dissolution occurred smoothly during the initial stage. Bright spots formed and accumulated until the entire shell became brighter, and no material was seen to leave the surface (images not shown). Gopal and Lee had observed that, in the two-phase system, attached vesicles only formed during monolayer compression at a temperature range between the critical temperature at which the two components become miscible and several degrees below the critical temperature. By direct comparison, using reduced temperature, the temperatures used here are below this range, yet vesiculation persists. We can think of two possible explanations, both suggested by the work of Lee and co-workers. DinPC

domain number

domain area before fold (µm2)

1 2 3 4 Total

96.3 25.5 18.4 78.1 218

Frames 1 to 2 89.3 18.0 13.4 53.5 174

7.3% 30% 27% 32% 20%

1 2 3 Total

33.2 87.5 51.7 172

Frames 3 to 4 23.9 52.7 22.8 99.4

28% 40% 56% 42%

domain area after fold (µm2)

area change percentagea

a Area change percentage ) (area before fold - area after fold)/area before fold × 100%.

+ PEG40S are maintaining a degree of miscibility similar to 16:0-18:1PG + Di16PC several degrees below their miscibility transition. This explanation suggests that some lipid exists with PEG40S in the LE phase. Our observation of vesiculation from the E phase supports this possibility since some lipid should be necessary for vesiculation (PEG40S forms micelles). Alternatively, for vesiculation to continue, C-phase domains may not be conferring sufficient rigidity upon LE-phase components in the domain boundary region. The difference in Πc of PEG40S and 16:0-18:1PG at RT (∼35 vs ∼55 mN/m) indicates that PEG40S forms a significantly less rigid monolayer compared to 16:0-18:1PG. Potentially, it is not possible for PEG40S to rigidify sufficiently to maintain it in the 2D state at high surface pressure. Secondary Stage: Folding. The collapse behavior following initial E-phase vesiculation, as described above, showed trends that were dependent on chain length and surface microstructure. Monolayers containing Di14PC + PEG40S at 5 °C and Di16PC + PEG40S at RT displayed similar behaviors. For bubbles that contained large and medium-sized domains (>2 µm), after the domains were packed together, dissolution occurred in discrete monolayer collapse events with smooth shrinkage between the events (Figure 2B and Movie 3, Supporting Information). In each collapse event that occurred in less than 300 ms (the time resolution of our camera), material was lost from the monolayer shell through an apparent fold. The size of the microbubble stepped down and deformed with each folding event. The line in the first image in Figure 2B delineates the apparent location of a virtually straight fold, which has run across four domains in view, resulting in the shrinkage of each domain along the same line (frames 1-2). This was followed by another fold parallel to the first, running through three domains, and possibly along the same line (frames 3-4). Interestingly, as can be seen in Figure 2B, the fold lines did not follow the domain edges. We tracked the amount of material lost from the domains for this bubble (Table 2) and others. Each folding event resulted in significant material loss. For example, in the second folding event shown in Figure 2B (frames 3-4), domain #3 lost ∼55% of its surface area to the fold. By observing the edge (rather than the top) of the bubble, it could be seen that the intervening bright E phase was also folding during these events, as evidenced by the growing bright material extending out from the bubble between domains (arrows). This material remained attached to the monolayer. These observations are strikingly similar to images shown for 16:0-18:1PG + Di16PC, in which wide fold structures correspond to the presence of large domains and the domain structure persisted in the folds.32 In monolayers of 16:0-18:1PG + Di16PC, the boundary between the intervening E phase and the

2998 Langmuir, Vol. 22, No. 7, 2006

C domains provided the nucleation site for folding at the C-phase Πc (near 72 mN/m). Here, we suspect that nucleation of folds occurred in the same region. (It is reported that a time resolution of