Calcein

ULTRASOUND-INDUCED CALCEIN RELEASE FROM eLIPOSOMES

JAMES R. LATTIN,

* WILLIAM G. PITT,

* DAVID M. BELNAP, and GHALEB A. HUSSEINI

*Chemical Engineering Department; Chemistry and Biochemistry Department, Brigham Young University, Provo, UT, USA;
and Chemical Engineering Department, American University of Sharjah, Sharjah, United Arab Emirates
(Received 24 April 2012; revised 28 July 2012; in final form 1 August 2012)
Abstract—Ultrasound is explored as a method of inducing the release of encapsulated materials from eLiposomes, defined as liposomes containing emulsion droplets. Emulsions were formed using perfluorohexane and perfluoro- pentane. eLiposomes were formed by folding interdigitated lipid sheets into closed vesicles around the emulsion droplets. Cryogenic transmission electron microscopy was used to verify droplet encapsulation. Self-quenched calcein was also encapsulated inside the vesicles. A fluorometer was used to measure baseline fluorescence, calcein release after ultrasound exposure, and total release from the vesicles. eLiposome samples released 3 to 5 times more of the encapsulated calcein than did controls when exposed to 20-kHz ultrasound. Calcein release increased with exposure time and intensity of ultrasound. eLiposomes with large (400 nm) droplets produced more calcein release than small (100 nm) droplets. These observations suggest that the emulsions are vaporized by ultrasound and that the Laplace pressure in the emulsions has an effect on droplet vaporization.
Key Words: Liposome, Emulsion, Perfluorocarbon, Ultrasound, Drug delivery.

INTRODUCTION
Liposomes (bilayered lipid vesicles) have proven to be effective and versatile drug carriers. The ability to change the properties of the lipid membrane by changing its composition adds to their versatility (Gregoriadis and Florence 1993; Lindner and Hossann 2010). Another key characteristic of liposomes is the ability to control their size (Gregoriadis and Florence 1993), which allows them to be used for ‘‘passive’’targeting to cancerous tumors (Seymour 1992). The chaotic and leaky endothe- lium of many cancerous tissues allows submicron parti- cles to extravasate into the tumor and accumulate in higher concentrations compared to their accumulation in normal tissues. This tendency of submicron particles to extravasate and accumulate in cancerous tissues is known as the enhanced permeability and retention (EPR) effect (Maeda 2001). Data seem to suggest that optimum vesicle sizes for the EPR effect range between 100 nm and 300 nm (Hobbs et al. 1998; Bae and Park 2011). Lipo- somes can be formed at specific sizes to take advantage of the EPR effect, making them particularly good candi- dates for delivering therapeutics to tumors. However, although passive targeting does improve selectivity to a target site, substantial amounts of the drug still accumu- late in other tissues (Bae and Park 2011).
Active targeting of drug delivery from liposomes could also increase the efficiency of their delivery by more specifically controlling the site of drug delivery after they have accumulated in tissues. Active targeting typi- cally refers to attaching targeting ligands to the surface of the liposomes (Maruyama et al. 1999). These ligands are recognized by a target cell type and aid in preferential attachment and sometimes induce uptake of the therapeu- tics by targeted cells (Lee and Low 1997; Bareford and Swaan 2007; Gabizon et al. 2010). Another form of target- ing is called actuated targeting and involves the use of external triggers. For example, ultrasound can be used to release drugs from carriers at a specific target site. Ultra- sound has the advantages of being noninvasive and allow- ing control of both the site and the timing of release (Husseini and Pitt 2008a, 2008b) and can be used in conjunction with active and passive targeting. Ultrasound can be focused on targets deep in the body without affecting other tissues, allowing the carriers to sequester their therapeutic content until induced to release the drug load at a specific site (Rahim et al. 2006).

Additionally, ultrasound increases vasculature and cell membrane permeability to facilitate drug uptake (Guzman et al. 2001; Prentice et al. 2005; Meijering et al. 2009; Lin et al. 2010). Ultrasound has been used to induce drug carriers such as microbubbles to permeabilize cells or to release their drug load at a specific location or both (Unger et al. 1998; Prentice et al. 2005; Liu et al. 2006). However, it is difficult to form stable gas bubbles at sizes sufficiently small to take advantage of the EPR effect. Micelles have also been investigated as ultrasound-sensitive drug carriers (Rapoport et al. 2004; Husseini and Pitt 2008b). Because of their versatility and other advantages, liposomes are also candidates for ultrasound-induced drug delivery (Schroeder et al. 2007; Klibanov et al. 2010; Evjen et al. 2011; Afadzi et al. 2012). However, liposomes are not inherently responsive to ultrasound. Ultrasonic drug release from conventional liposomes seems to rely on the presence of gas bubbles to cause cavitation and is dependent on lipid composition (Schroeder et al. 2007; Schroeder et al. 2009; Evjen et al. 2011). A relatively new field is that of echogenic lipo- somes. These liposomal drug carriers are typically de- signed to contain small amounts of gas in order to increase ultrasound sensitivity (Huang and MacDonald 2004; Huang 2008; Kopechek et al. 2008). Echogenic liposomes have demonstrated ultrasound-induced release of hydrophilic compounds from liposomes (Kopechek et al. 2008) and have shown the potential to be used as diagnostic agents (Kopechek et al. 2011).
The goal of the current work is to develop a submi- cron-sized liposomal drug carrier that is responsive to low-intensity ultrasound without the requirement of nearby gas bubbles that are too large to extravasate. The proposed drug carrier is called an eLiposome, defined as a liposome containing vaporizable emulsion droplets (Lattin et al. 2012). The nanosized emulsion droplets are formed from perfluorocarbon liquids with high vapor pressures. During the low-pressure phase of an ultrasound wave, the local pressure may drop below this high vapor pressure, allowing the formation and expansion of a vapor phase (Sheeran et al. 2011). As this vapor phase forms and expands, it stretches and disrupts the liposomal membrane, causing local drug release. Emulsion droplets can be formed on a much smaller scale than bubbles, allowing the eLiposome to be small enough to extravasate via the EPR effect while also introducing ultrasound sensitivity to this liposome- like drug carrier. This article presents the release of encapsulated material from eLiposomes when exposed to 20-kHz ultrasound. Calcein, a model drug, was encap- sulated inside the eLiposomes, and fluorescence spectros- copy was used to quantify ultrasound-mediated release. Ultrasound intensities and durations of exposure were varied.

Materials
Perfluorohexane (PFC6) was purchased from Sigma-Aldrich (St. Louis, MO, USA). Perfluoropentane (PFC5) was purchased from SynQuest Labs (Alachua, FL, USA). Dipalmitoyl phosphatidyl choline (DPPC) dis- solved in chloroform was purchased from Avanti Polar Lipids (Alabaster, AL, USA). Calcein was purchased from MP Biomedicals (Solon, Ohio, USA). All water used was double distilled.

Emulsion preparation
Emulsions were made using PFC5 and PFC6. A thin film of DPPC was deposited on a round-bottomed flask by evaporation of the chloroform. The lipid film was then hydrated at 30 mg/mL by adding water and stirring at 50 C. Then 167 mL of this hydrated DPPC suspension (5 mg DPPC) was added to a vial along with 0.2 g of PFC and 1.5 mL of water. The water was degassed prior to emulsion formation. Degassing could not be performed after emulsion preparation without destroying the emul- sion. The flask was placed in an ice bath and allowed to cool. The emulsion was formed by sonicating the contents with a 20-kHz ultrasound probe 3 mm in diam- eter (Vibra-Cell VCX400, Sonics and Materials, Newton, CT). Ultrasound intensity was previously calibrated in degassed and deionized water using a low-frequency hydrophone (Model 8103, Bruel & Kjaer, Naerum, Denmark). The hydrophone was positioned 3 mm from the transducer and connected to an oscilloscope. This calibration was performed in a 100-L water tank lined with corrugated rho C rubber to minimize potential inter- ference at the hydrophone tip resulting from standing waves. The hydrophone voltage signal was converted into intensity (W/cm ) values using the factory calibra- tion parameters. Large emulsion droplets were prepared by sonicating for 1 min at 0.5 W/cm . Small emulsion droplets were formed by sonicating for 1 min at 1 W/cm . The droplet size was further reduced and controlled by extrusion through a 100-nm polycarbonate membrane; the PFC emulsion was heated to 50 C and passed through the membrane 15 times using an Avanti Mini Extruder (Alabaster, AL, USA). Droplet size was verified by dynamic light scattering on a Brookhaven 90Plus Particle Sizer (Brookhaven Instruments, Holtsville, NY, USA). Large emulsion droplets were considered suitable to form eLiposomes if the median diameter ranged between 400 and 500 nm. Small emulsion droplets were used if the median diameter was between 90 and 120 nm.
Ultrasonic vaporization of emulsion droplets was verified by applying ultrasound to the emulsion sand- wiched between a microscope slide and a cover slip at room temperature (24 C). Then 100 mL of emulsion was diluted in 1 mL of water. The diluted emulsion was viewed with an Olympus IX70 microscope using a 340 objective. The slide was then inverted and exposed to 1 W/cm ultrasound for 5 s by applying a small amount of ultrasound gel to the slide and positioning the 20 kHz transducer directly into the gel. The slide was again imaged after ultrasound exposure. Control experi- ments were performed using plain water and a DPPC lipid solution to verify that bubbles were not formed in the absence of PFC emulsion.

eLiposome Preparation
eLiposomes were formed using a lipid sheet refold- ing technique (Kisak et al. 2004; Boyer and Zasadzinski 2007). Small (30 to 70 nm) DPPC liposomes were prepared by thin film hydration followed by 20-kHz soni- cation for 15 min at 1 W/cm . Size was verified by dynamic light scattering and, if necessary, the sample was sonicated for an additional 15 min. The resulting solution of small DPPC vesicles was translucent. Inter- digitated DPPC sheets were formed from these small vesicles by adding ethanol dropwise to a total concentra- tion of 3 M. Formation of lipid sheets was manifested by a large increase in viscosity and the loss of translucence. After sheet formation, excess ethanol was removed from the sheet suspension by centrifugation and washing. The sheets were suspended and diluted in 50 mL of water at room temperature and centrifuged at 18003 g for 3 min, resulting in a large pellet of DPPC sheets. The alcohol-rich supernatant was removed, and the pellet was resuspended in water and centrifuged again. The pellet was again suspended in a small volume of water in order to transfer the sheets to microcentrifuge tubes, with 10 mg of DPPC allotted to each tube. The extra water was removed through 1 more cycle of centrifuga- tion, and 0.2 mL of emulsion was added to 10 mg of inter- digitated sheets along with 0.2 mL of water (for imaging by transmission electron microscopy [TEM]) or a 30-mM calcein solution (for release experiments). The sheet solu- tion was repipetted and briefly vortexed to ensure complete mixing. The solution was then heated to 50 C and stirred with a magnetic stir bar for 30 min, allowing the sheets to fold back into vesicles and trapping some nanoemulsion droplets inside (Lattin et al. 2012). The size distribution of the resulting eLiposomes was controlled by extrusion through an 800-nm polycarbonate filter at 50 C. Empty eLiposomes, defined as liposomes refolded from interdigitated sheets in the absence of emulsion, were prepared as negative controls.

Transmission electron microscopy imaging
The eLiposomes were imaged using cryogenic TEM (cryoTEM). eLiposome samples were pipetted onto a holey-carbon-coated copper grid. A Vitrobot (FEI, Hillsboro, OR, USA) was used to blot the grid with filter paper and plunge it into liquid ethane for freezing. Frozen grids were stored in liquid nitrogen. A 626 cryoholder (Gatan, Pleasanton, CA, USA) maintained the samples at approximately 2180 C in the microscope. The micro- scope stage was rotated and the sample was viewed at 245 , 0 , and 145 to verify encapsulation of the emul- sion droplets (Lattin et al. 2012). Images were recorded at 300 kVon a Tecnai F30 transmission electron microscope (FEI) using a Gatan 1024 3 1024 CCD camera. The objective lens was underfocused by several micrometers to improve contrast.

Calcein release
Calcein-containing samples were prepared for quan- tification of release. During vesicle formation, calcein was encapsulated inside of the eLiposomes at a concentration of approximately 15 mM. At this concentration, the cal- cein was self-quenched. The external calcein concentra- tion was reduced by allowing the sample to settle at the bottom of the microcentrifuge tube for a few hours. Because of their relatively high density, the vesicles settled to the bottom of the tube in a thick gel-like phase. The top vesicle-free layer was removed, and the samples were resuspended in an NaCl solution with the osmolarity matched to the calcein solution inside the eLiposomes. This created samples with concentrated calcein inside the eLiposomes and dilute calcein on the outside. A corre- lation of calcein fluorescence versus concentration was prepared using a QuantaMaster fluorometer (Photon Technology International, Birmingham, NJ, USA) with excitation and emission wavelengths of 488 nm and 525 nm, respectively. At concentrations below 7.5 mM calcein, the plot of fluorescence intensity versus calcein concen- tration is linear. The eLiposome samples were diluted by adding 20 mL of sample to a disposable UV/VIS- range cuvette with a 10-mm path length (Fisher Scientific, Pittsburgh, PA, USA). Then 2 mL of NaCl solution was added to the cuvette to achieve a target external calcein concentration of 1 to 5 mM in order to operate in the linear region of the concentration curve for calcein. Standards were prepared at 1 mM and 5 mM to test the fluorescence emission (counts per second) expected for the target concentration range in order to verify that experiments were being run within these concentrations.
Baseline fluorescence data were collected for 10 s at 4 data points per s. The cuvette was then removed from the fluorometer and 20-kHz ultrasound was applied using a 3- mm transducer (Sonics and Materials, Newton, CT, USA) inserted directly into the cuvette. As concentrated (self- quenched) calcein was released from the interior of the eLiposomes into the surrounding solution, it was diluted below its self-quenching concentration. Fluorescence was again measured after sonication. Finally, 25 mL of 5% Triton X-100 was added to lyse any remaining lipo- somes, and a final fluorescence was measured. All exper- iments were conducted at 24 C. Percent release of calcein was determined using the following equation:

an 800 nm eLiposome with 3 distinct 100 to 150-nm emul- sion droplets. Figure 1b shows an example of an 800-nm eLiposome with a single 470-nm emulsion droplet as well as 2 smaller droplets.

In order to verify acoustically produced vaporization of emulsion droplets, large PFC5 emulsions were placed on a microscope slide, were covered with a thin glass

where fi is initial fluorescence, fUS is the fluorescent signal after sonication, and ffull is the fluorescence after complete calcein release using Triton X-100. Release experiments were performed on eLiposome samples and on empty control liposome samples. Experiments were also performed on empty control samples mixed with emulsion droplets after the formation of the control vesicles, such that the emulsion droplets were external to the liposomes instead of internal.

RESULTS
CryoTEM images of eLiposomes verify that PFC6 emulsion droplets were encapsulated inside the lipid vesi- cles using the sheet refolding technique. Figure 1 shows examples of eLiposomes containing emulsion droplets. The interdigitated sheet refolding technique was used to encapsulate both small and large emulsion droplets. The dark circular structures are the perfluorocarbon droplets. These droplets are observed within a DPPC membrane layer. The microscope stage was rotated to 245 , 0 , and 145 to verify encapsulation of the emulsion droplets within the lipid vesicle. The emulsion droplets were observed within the membrane boundary as the micro- scope stage was rotated, demonstrating that they were encapsulated inside of the eLiposome (data not shown) (Lattin et al. 2012). Figure 1 demonstrates that both small and large droplets can be encapsulated inside the newly formed DPPC vesicles. Figure 1a shows an example of coverslip, and were exposed to 1 W/cm ultrasound for 5 s. Prior to ultrasound exposure only the largest droplets (approximately 3 to 4 mm) are visible (Fig. 2a). After ultra- sound exposure, many larger structures can be observed that have the appearance of gas bubbles (Fig. 2b). These newly formed bubbles vary in size and can be as large as 20 mm. The study included 2 control experiments involving a DPPC solution and plain water. No bubbles were observed before or after ultrasound exposure in either of the control samples (data not shown).
When exposed to ultrasound, eLiposome samples demonstrated an increase in calcein release compared to controls. Figure 3 shows the percent released after 100 ms of exposure to 20 kHz ultrasound (n 53, and error bars represent 1 standard deviation). Ultrasound intensity was varied between 0.5 W/cm and 5 W/cm . eLiposomes were prepared with large (400 nm) and small (100 nm) emulsion droplets. The amount of release increased as ultrasound intensity increased. After 100 ms of ultrasound exposure, PFC5 eLiposomes released approximately 3 to 4 times more calcein than controls at each intensity (Fig. 3a). For example at 5 W/cm , eLiposomes with large PFC5 emulsion droplets released 39% and eLiposomes containing small PFC5 emulsion droplets released 31%, while control samples released only 10% of the encapsu- lated calcein. PFC6 eLiposomes also released more of the encapsulated calcein than control samples. Figure 3b shows calcein release from PFC6 eLiposomes after 100 ms of insonation. Although this release was slightly less than that observed in PFC5 eLiposomes, PFC6 eLipo- somes also released up to 3 times more calcein than control samples. After 100 ms at 5 W/cm , eLiposomes with 100-nm PFC6 emulsions released 22% of the avail- able calcein. The difference in calcein release between controls and eLiposomes was statistically significant (Student t-test, p ,0.05) for both PFC5 and PFC6 eLipo- somes with both large and small emulsions. Negative controls included both empty vesicles and empty vesicles with emulsion droplets added to the exterior. All of these negative controls released less of their encapsulated cal- cein than did the eLiposome samples. At 100 ms of ultra- sound exposure there was no significant difference (p .0.05) between empty control liposomes and empty liposomes with emulsion droplets added to the outside solution, regardless of emulsion size or material (Fig. 3).

There was, however, more variability in samples with emulsion droplets added to the outside of the vesicles; empty vesicles had an average standard deviation of 3% compared to a standard deviation of 5% when emulsion droplets were added to the exterior solution (n 53). There was no significant difference (p .0.05) between release from empty vesicles with small exterior droplets and release from empty vesicles with large exterior droplets.
The amount of calcein release also increased with increasing time of ultrasound exposure, as shown in Figure 4, which plots the release as the time of ultrasound exposure was varied from 100 ms to 10 s at 1 W/cm . Each point is the mean of 3 measurements. After 10 s, the PFC5 eLiposomes with 400-nm emulsion droplets had released 94% of the encapsulated calcein compared to only a 20% release from control samples. The same trend of increased calcein release with increased time was observed in all samples. PFC6 eLiposomes with 400-nm droplets released 47% of encapsulated calcein after 10 s. This was much less than PFC5 eLiposomes with large droplets, but more than twice as much release as the controls.
The size of the emulsion droplets had only a small effect at 100 ms (Fig. 3), but this effect became more pronounced with PFC5 eLiposomes at longer ultrasound exposures. PFC5 eLiposomes with 400-nm emulsion droplets released approximately 94% of encapsulated calcein after 10 s of 1 W/cm ultrasound, whereas PFC5 eLiposomes with 100-nm droplets released only 47% after 10 s. The difference in calcein release with droplet size was less pronounced for PFC6 eLiposomes; eLipo- somes containing 400-nm droplets released 45% of the encapsulated calcein, and eLiposomes with 100-nm PFC6 droplets released 34%. All eLiposome samples released more calcein than did negative controls, which released 20% of the encapsulated calcein after 10 s at 1

W/cm . The difference in calcein release between controls and eLiposomes was statistically significant for PFC5 and PFC6 eLiposomes at each time point for both large and small emulsion droplets (p ,0.05). Similar to the results shown in Figure 3 (various intensities after 100 ms of ultrasound exposure), control samples with external emul- sion droplets released the same average amount as empty negative controls but showed more variability. Sham experiments were also performed in which eLiposomes with large emulsion droplets were submitted to the entire release procedure, including inserting the probe into the cuvette for the same amount of time as in active experi- ments. These sham experiments showed no evidence of leaking or eLiposome damage resulting from the experi- mental procedure (Fig. 4).
Ultrasound intensity was also varied at 10 s of expo- sure (Table 1). In most cases, there was significantly more release from eLiposomes exposed to 1 W/cm than from those exposed to 0.5 W/cm for 10 s. For example, PFC5 eLiposomes with 400-nm droplets released 75% of encapsulated calcein at 0.5 W/cm and 94% at 1 W/cm . Similarly, PFC6 eLiposomes with 400-nm droplets released 33% of encapsulated calcein at 0.5 W/cm and 45% at 1 W/cm . These differences were statistically significant in all eLiposome samples except PFC6 eLipo- somes with small droplets (p ,0.05). However, when ultra- sound intensity was increased from 1 W/cm to 2 W/cm , there was no significant additional change in the amount of calcein released from any of the eLiposome samples. In contrast, the difference in calcein release by controls was generally not significant when ultrasound was increased from 0.5 W/cm to 1 W/cm . However, all control samples (both empty and with external emulsion) showed a statistically significant increase in calcein release after 10 s of ultrasound when the intensity was increased from 1 W/cm to 2 W/cm . The temperature of the solution was measured after sonication. The temperatures of samples exposed to 2 W/cm for 10 s increased by an average of approximately 4 C and those of samples exposed to 2 W/cm for 5 s increased by approximately 2 C. Increased temperature was not measurable in other samples at lower intensities and times. Because the resulting temper- atures were still well below the transition temperature for DPPC and below the temperature at which the nanoe- mulsions were extruded (50 C), it is unlikely that the increase in temperature was responsible for the increase in calcein release from eLiposomes or from control samples.

DISCUSSION
eLiposomes were successfully formed by adding emulsion droplets to interdigitated DPPC sheets and re- folding the lipid sheets into vesicles in the presence of emulsion. This method was able to encapsulate both large (400 nm) and small (100 nm) emulsion droplets, as demonstrated by CryoTEM. Rotating the microscope stage verified that the droplets were inside the vesicles. PFC5 and PFC6 were chosen for emulsion droplets because of their relatively high vapor pressures. Addition- ally, both of these perfluorocarbons have very low solu- bility in water, enabling nanodroplets to persist for an extended time in an aqueous environment. Finally, PFC5 and PFC6 have low toxicity (Correas et al. 2001; Nieuwoudt et al. 2009), making them good candidates for use in a drug delivery system. The ability of ultrasound to vaporize PFC emulsion droplets was verified by visual- izing emulsion droplets before and after ultrasound expo- sure (Fig. 2). Images taken after ultrasound exposure revealed that large bubbles had formed. These resulting bubbles were larger than expected and persisted at room temperature. The larger size could be due to bubbles’ colliding and combining or due to dissolved nitrogen and oxygen accumulating in the PFC bubble during and perhaps after insonation.

Although most medical ultrasound is of higher frequency, ultrasound was applied at 20 kHz in order to employ an acoustic system that could be applied within an optical cuvette. The work presented in this article is intended to demonstrate the greater sensitivity of eLipo- somes to ultrasound, as compared to conventional lipo- somes, in order to explore the potential of liposomes to be disrupted by acoustic emulsion droplet vaporization at short exposure times and relatively low ultrasound intensity. Although ultrasound from a probe sonicator would not be used in a clinical setting (because of focus and potential safety limitations), investigation of the ultra- sound parameters provides insight into the ability to use eLiposomes at higher frequencies in a clinical setting or at the same frequency in a biological laboratory. The short exposure times and lower intensities chosen provide ultra- sound exposures that would be comparable to higher frequency ultrasound administered in short pulses over a longer period of time. Acoustic droplet vaporization of perfluorocarbons has been reported by others at higher frequencies and similar mechanical indexes (Fabiilli et al. 2010; Rapoport et al. 2011; Sheeran et al. 2011). The mechanical indexes at the lowest 3 intensities used (0.5, 1, and 2 W/cm ) were 0.63, 1.22, and 1.73, respec- tively, which may be clinically acceptable for short pulses to tumors. Furthermore, 20-kHz ultrasound at similar mechanical indexes has been used for drug delivery in cancer therapy in animal models (Larina et al. 2005; Staples et al. 2009). The eLiposome samples demon- strated increased calcein release compared to conven- tional liposomes (containing only calcein) and compared to conventional liposomes with external emul- sions, suggesting that the encapsulated emulsion droplets add an ultrasound-sensitive element to the liposomes. The proposed mechanism for this increased sensitivity to ultrasound is the ability of ultrasound to vaporize the emulsion droplets (Kripfgans et al. 2000; Giesecke and Hynynen 2003; Rapoport et al. 2011; Singh et al. 2012). When the local pressure inside of the emulsion droplet falls below the vapor pressure of the emulsion there is a thermodynamic potential to vaporize. The expanding vapor phase stretches and disrupts the surrounding membrane and releases the encapsulated material because lipid bilayers can sustain only about a 3% expansion (Evans et al. 1976; Netz and Schick 1996). Two possible representations of this are illustrated in Figure 5. As the vapor phase expands, it may stretch the membrane, creating small rips and expelling the encapsulated mate- rial (Fig. 5a). This scenario is perhaps most likely with small droplets relative to eLiposome size and with a liquid having a lower vapor pressure. Perfluorocarbons with lower vapor pressures may be more difficult to vaporize and will collapse back into liquid droplets when no longer exposed to the low-pressure phase of an ultrasound wave.

Fig. 5. Proposed mechanisms of ultrasound-induced release from eLiposomes. The small ovals represent calcein or encapsulated drugs. In each panel the medium-sized circle inside the liposome represents a liquid perfluorocarbon emul- sion droplet. When exposed to the low pressure phase of an ultrasound wave, the droplet vaporizes and expands, disrupt- ing the bilayered membrane and releasing the contents from the eLiposome. This expansion may stretch and rip the membrane (a), or it may more completely destroy the liposome by fragmenting the bilayered membrane into small
sections (b).

In this scenario, the gaps formed in the stretching membrane could potentially close during each cycle of ultrasound and following ultrasound exposure. Alterna- tively, if the expansion is more violent and/or if the volume of perfluorocarbon is more substantial, the membrane could be fragmented by the expanding vapor phase and associated cavitation events (Fig. 5b).
Remarkably, in the absence of ultrasound the eLipo- some membranes appear to be sufficiently robust to completely sequester their cargo of calcein through various laboratory manipulations. The sham experiments showed no leakage (Fig. 4).
The eLiposomes were most sensitive to ultrasound when large emulsion droplets were encapsulated. This is most likely due to the greater Laplace pressure imposed on the droplets as their sizes decrease. The pressure on the nanodroplet interior is the sum of the environmental pres- sure and the Laplace pressure, or:

that it would boil at body temperature (37 C). However, the additional Laplace pressure imposed on PFC5 emul- sion droplets effectively raises the boiling point of small droplets above their normal boiling point, allowing them to remain in the liquid phase well above biological temperatures (Giesecke and Hynynen 2003; Sheeran et al. 2011), creating a PFC5 emulsion droplet that is meta-stable until disrupted by ultrasound. For example, when heated to 50 C for extrusion in our lab, emulsion droplets persisted despite being heated above the normal boiling point of PFC5. A slight foam layer sometimes formed after extrusion. The presence of emulsion drop- lets after heating and extrusion was verified by adding an emulsion to the top layer of a ‘‘sucrose cushion.’’ The 1 M sucrose layer had a density much greater than that of perfluorocarbon bubbles, slightly greater than that of an aqueous suspension of liposomes, but less than that of liquid perfluorocarbons. The collection of a pellet at the bottom of the sucrose layer verified the preservation of emulsion droplets. Persistence of nano- sized emulsion droplets was further confirmed by TEM.

where Pinside is the pressure inside of the droplet, PNis the local pressure of the continuous water phase, g is the interfacial tension of the PFC/water interface, and r is the radius of the droplet. The increased pressure on the droplet adds to the ultrasound intensity required to vaporize the droplet. Because Laplace pressure is inversely proportional to radius, the amount of pressure on the droplet increases as droplet diameter decreases. The Laplace pressure across the interface of the nanoe- mulsions can be estimated by using the interfacial tension reported for other perfluorocarbon emulsions stabilized by phosphocholines (Kabalnov et al. 1995). Interfacial tension for a perfluorocarbon droplet in water stabilized with DPPC can be estimated to be about 3.5 mN/m. This predicts a Laplace pressure of 140 kPa for 100-nm droplets and 35 kPa for 400-nm droplets. It may be noted that PFC5 has a normal boiling point of 29 C, suggesting

In the case of a meta-stable emulsion formed from a per- fluorocarbon with a boiling point that is lower than bio- logic temperatures, droplet vaporization may be irreversible at biologic temperatures due to the decrease of the Laplace pressure as the bubble radius increases upon gas expansion. This irreversible phase change would create a persistent oscillating bubble that would continue to disrupt and shear the lipid membrane (Rapoport et al. 2011).
The increase in Laplace pressure with decreased droplet size results in a higher required ultrasound inten- sity to vaporize smaller droplets. As mentioned previously, both PFC5 and PFC6 eLiposomes with large emulsion droplets released more calcein than those with small drop- lets. In this study, the effect of droplet size can be best observed with PFC5 eLiposomes: eLiposomes with large emulsion droplets released approximately twice as much calcein as those with small droplets after 10 s of ultrasound exposure at all 3 intensities tested (Table 1). After 100 ms of ultrasound exposure, the difference in calcein release from eLiposomes with large PFC5 emulsion droplets was statistically significant (p ,0.05) compared to that of eLiposomes with small droplets at intensities of 2 W/ cm or greater. At 1 W/cm , these differences were signif- icant for samples exposed to ultrasound for 1 s or longer. In PFC6 samples exposed to 100 ms of ultrasound, the differ- ence in calcein release from eLiposomes with large emul- sion droplets was statistically significant compared to that of eLiposomes with small droplets at intensities of 3 W/cm or greater. At 1 W/cm , these differences were statistically significant for samples exposed to ultrasound for 5 s or longer.
There was also a noticeable difference between eLi- posomes formed with PFC5 and PFC6. This difference is most likely due to the difference in vapor pressures. Exper- iments were conducted at 24 C. At this temperature, the vapor pressure of PFC6 is 28 kPa and the vapor pressure of PFC5 is 84 kPa. Because of the higher vapor pressure of PFC5, less acoustic amplitude is required to drop the local pressure below the PFC5 vapor pressure. At only 0.5 W/cm , the ultrasound imposes a negative pressure of 122 kPa, sufficient to overcome the difference in the atmospheric pressure and the vapor pressure of PFC5 or PFC6. This amplitude is also sufficient to overcome the Laplace pressure on 400-nm droplets. For 100-nm PFC5 droplets, 1 W/cm is required to overcome the Laplace pressure, with ultrasound at that intensity imposing 173 kPa of negative pressure. Higher intensity ultrasound (2 W/cm ) is required to provide 245 kPa of negative pres- sure in order to overcome the Laplace pressure and the lower vapor pressure for 100-nm PFC6 droplets. The expected pattern of PFC5 eLiposomes releasing more cal- cein than PFC6 eLiposomes was consistent for all analo- gous samples. These differences were statistically significant in a Student t-test (p ,0.05) for all intensities and times reported except for eLiposomes with small droplets after 10 s of ultrasound exposure at 0.5 W/cm (Table 1).
Our eLiposomes released more calcein than the negative controls of liposomes without any emulsions or liposomes with external emulsions. This is consistent with the hypothesis that increased release was due to internal emulsion droplets’expanding and disrupting the membrane. Short exposure times (100 ms) led to increased release from eLiposomes compared to controls, which is also consistent with this hypothesis. In order to vaporize a liquid emulsion droplet with ultrasound, the applied negative pressure must lower the environmental pressure below the liquid vapor pressure and overcome the Laplace pressure imposed on the droplet. When these conditions are not met, the liquid droplet will remain in the liquid phase. After imposing the necessary pressure, a vapor phase should nucleate and expand during phases of low pressure and rupture the eLiposome membrane. The required ultrasound intensity for vaporization and time for nucleation should result in some kind of inten- sity threshold for calcein release from eLiposomes. However, obvious thresholds were not observed with respect to time or intensity, despite operating at ultra- sound parameters above and below those calculated to overcome both the Laplace pressure on the droplets and the difference between the local pressure and the droplet vapor pressure. This lack of obvious thresholds is most likely due to the distribution of sizes in both emulsion droplets and eLiposomes. The distribution of emulsion sizes results in differing Laplace pressures being imposed from droplet to droplet. Some emulsion droplets will therefore require higher or lower negative pressure in order to vaporize, and this distribution may be responsible for the lack of an intensity threshold for calcein release. Furthermore, nucleation probably does not occur instantly but may be a stochastic process that is related to the degree of subpressurization of the emul- sion. The experimental data indicate that nucleation occurs, but there may be a distribution in nucleation times that may smear out a theoretically sharp threshold in experimental data.
At longer exposures and higher intensities, control samples began to release more calcein, and the gap in percent of calcein released between eLiposomes and negative controls became narrower. However, these higher intensities of ultrasound are less desirable for clin- ical use. eLiposomes have the advantage of releasing encapsulated loads at low ultrasound intensities and rela- tively short exposure times. Low-intensity exposure provides leverage to increase the amount of drug that can be released while limiting unwanted heating and cavitation in adjacent tissues. Furthermore, eLiposomes with large PFC5 droplets approached 100% release at reasonable ultrasound parameters. This release was much higher than that observed in control samples, even at longer exposure times and with higher ultrasound intensities. However, in order to take best advantage of the EPR, eLiposomes have to be reduced to sizes closer to 250 nm. This will require the use of smaller (100 to 125 nm) emulsion droplets. At these sizes, the additional Laplace pressure may limit the ability to release 100% of encapsulated drug from PFC5 or PFC6 eLiposomes at reasonable ultrasound intensities. Ideally, a 250-nm eLi- posome with small emulsion droplets could accumulate in cancer tumors or even in cells followed by an ultrasound-induced release of the entire encapsulated drug load. In order to approach this goal with eLipo- somes, more efficient methods of inducing emulsion va- porization or other materials may be required, such as using lower-boiling-point perfluorocarbons or mixtures thereof with PFC5.
Finally, we caution readers that it is probable that applying 20-kHz ultrasound (which has a relatively long wavelength of 7.5 cm) within a small plastic chamber created standing waves. Constructive and destructive interference caused by these standing waves may have resulted in local regions of higher and lower pressures than the reported calibrated values. Lower pres- sure caused by destructive interference could have reduced the amount of calcein release expected at any specific intensity, while constructive interference may have caused local regions of higher intensity. Because of mixing in the cuvette, it is likely that eLiposomes in regions of lower intensity would have been quickly con- vected to regions of higher intensity. Thus the results pre- sented herein represent the average release over the entire volume of a cuvette, and there may have been local regions of more intense insonation than the nominal intensity values listed in this report. Because the amount of release from the eLiposomes is not a linear function of intensity or time (Figs. 2, 3), extrapolation of these data to other experimental or clinical systems should be done with caution.

CONCLUSION
Liposomes formed from DPPC were made more responsive to ultrasound by encapsulating emulsion drop- lets inside of them. This was accomplished by forming lipid sheets, adding a suspension of emulsion droplets, and refolding the sheets into vesicles, encapsulating emul- sion droplets in the process. The resulting eLiposomes demonstrated excellent sequestration of payload in the absence of ultrasound and yet were sensitive to short pulses of low-intensity ultrasound. Upon insonation, severalfold more calcein was released from eLiposomes than from controls. As expected by thermodynamic considerations, the small emulsion droplets were not as efficient as the large droplets, and the PFC6 was not as efficient as the PFC5 in activating calcein release. These data point to the use of large PFC5 emulsions in eLipo- somes, but the need to construct eLiposomes small enough to be extravasated suggests that an optimized intermediate size would be most clinically relevant for drug delivery applications to tumors exhibiting the EPR effect.

REFERENCES
Afadzi M, Davies CD, Hansen YH, Johansen T, Standal OK, Hansen R, Masoy SE, Nilssen EA, Angelsen B. Effect of ultrasound parameters

on the release of liposomal calcein. Ultrasound Med Biol 2012;38: 476–486.
Bae YH, Park K. Targeted drug delivery to tumors: Myths, reality and possibility. J Control Release 2011;153:198 –205.
Bareford LA, Swaan PW. Endocytic mechanisms for targeted drug delivery. Adv Drug Deliver Rev 2007;59:748 –758.
Boyer C, Zasadzinski JA. Multiple lipid compartments slow vesicle contents release in lipases and serum. ACS Nano 2007;1: 176–182.
Correas JM, Meuter AR, Singlas E, Kessler DR, Worah D, Quay SC. Human pharmacokinetics of a perfluorocarbon ultrasound contrast agent evaluated with gas chromatography. Ultrasound Med Biol 2001;27:565 –570.
Evans EA, Waugh R, Melnik L. Elastic area compressibility modulus of red-cell membrane. Biophys J 1976;16:585 –595.
Evjen TJ, Nilssen EA, Barnert S, Schubert R, Brandl M, Fossheim SL. Ultrasound-mediated destabilization and drug release from lipo- somes comprising dioleoylphosphatidylethanolamine. Eur J Pharm Sci 2011;42:380 –386.
Fabiilli ML, Haworth KJ, Sebastian IE, Kripfgans OD, Carson PL, Fowlkes JB. Delivery of chlorambucil using an acoustically trig- gered perfluoropentane emulsion. Ultrasound Med Biol 2010;36: 1364–1375.
Gabizon A, Tzemach D, Gorin J, Mak L, Amitay Y, Shmeeda H, Zalipsky S. Improved therapeutic activity of folate-targeted lipo- somal doxorubicin in folate receptor-expressing tumor models. Cancer Chemother Pharmacol 2010;66:43–52.
Giesecke T, Hynynen K. Ultrasound-mediated cavitation thresholds of liquid perfluorocarbon droplets in vitro. Ultrasound Med Biol 2003;29:1359 –1365.
Gregoriadis G, Florence AT. Liposomes in drug delivery: Clinical, diag- nostic and ophthalmic potential. drugs 1993;45:15 –28.
Guzman HR, Nguyen DX, Khan S, Prausnitz MR. Ultrasound-mediated disruption of cell membranes. I. Quantification of molecular uptake and cell viability. J Acoust Soc Am 2001;110:588 –596.
Hobbs SK, Monsky WL, Yuan F, Roberts WG, Griffith L, Torchilin VP, Jain RK. Regulation of transport pathways in tumor vessels: Role of tumor type and microenvironment. Proc Natl Acad Sci U S A 1998; 95:4607–4612.
Huang SL. Liposomes in ultrasonic drug and gene delivery. Adv Drug Deliver Rev 2008;60:1167 –1176.
Huang SL, MacDonald RC. Acoustically active liposomes for drug encapsulation and ultrasound-triggered release. BBA Bio- membranes 2004;1665:134–141.
Husseini GA, Pitt WG. Micelles and nanoparticles for ultrasonic drug
and gene delivery. Adv Drug Deliver Rev 2008a;60:1137 –1152. Husseini GA, Pitt WG. The use of ultrasound and micelles in cancer
treatment. J Nanosci Nanotechnol 2008b;8:2205 –2215.
Kabalnov A, Weers J, Arlauskas R, Tarara T. Phospholipids as emulsion stabilizers. 1. Interfacial tensions. Langmuir 1995;11:2966 –2974.
Kisak ET, Coldren B, Evans CA, Boyer C, Zasadzinski JA. The veso- some: A multicompartment drug delivery vehicle. Curr Med Chem 2004;11:199 –219.
Klibanov AL, Shevchenko TI, Raju BI, Seip R, Chin CT. Ultrasound- triggered release of materials entrapped in microbubble-liposome constructs: A tool for targeted drug delivery. J Control Release 2010;148:13 –17.
Kopechek JA, Abruzzo TM, Wang B, Chrzanowski SM, Smith DAB, Kee PH, Huang S, Collier JH, McPherson DD, Holland CK. Ultra- sound-mediated release of hydrophilic and lipophilic agents from echogenic liposomes. J Ultrasound Med 2008;27:1597 –1606.
Kopechek JA, Haworth KJ, Raymond JL, Mast TD, Perrin SR, Klegerman ME, Huang SL, Porter TM, McPherson DD, Holland CK. Acoustic characterization of echogenic liposomes: Frequency-dependent attenuation and backscatter. J Acoust Soc Am 2011;130:3472–3481.
Kripfgans OD, Fowlkes JB, Miller DL, Eldevik OP, Carson PL. Acoustic droplet vaporization for therapeutic and diagnostic applications. Ultrasound Med Biol 2000;26:1177 –1189.
Larina IV, Evers BM, Ashitkov TV, Bartels C, Larin KV, Esenaliev RO. Enhancement of drug delivery in tumors by using interaction of nanoparticles with ultrasound radiation. Technol Cancer Res Treat 2005;4:217 –226.
Lattin JR, Belnap DM, Pitt WG. Formation of eLiposomes as a drug delivery vehicle. Coll Surf B: Biointerfac 2012;89:93 –100.
Lee RJ, Low PS. Folate-targeted liposomes for drug delivery. J Liposome Res 1997;7:455–466.
Lin CY, Liu TM, Chen CY, Huang YL, Huang WK, Sun CK, Chang FH, Lin WL. Quantitative and qualitative investigation into the impact of focused ultrasound with microbubbles on the triggered release of nanoparticles from vasculature in mouse tumors. J Control Release 2010;146:291 –298.
Lindner LH, Hossann M. Factors affecting drug release from liposomes. Curr Opin Drug Disc 2010;13:111 –123.
Liu YY, Miyoshi H, Nakamura M. Encapsulated ultrasound microbub- bles: Therapeutic application in drug/gene delivery. J Control Release 2006;114:89–99.
Maeda H. The enhanced permeability and retention (EPR) effect in tumor vasculature: The key role of tumor-selective macromolecular drug targeting. In: Weber G, (ed). Advances in Enzyme Regulation, Vol. 41. Oxford: Pergamon-Elsevier Science; 2001. p. 189–207.
Maruyama K, Ishida O, Takizawa T, Moribe K. Possibility of active tar- geting to tumor tissues with liposomes. Adv Drug Deliver Rev 1999; 40:89–102.
Meijering BDM, Juffermans LJM, van Wamel A, Henning RH, Zuhorn IS, Emmer M, Versteilen AMG, Paulus WJ, van Gilst WH, Kooiman K, de Jong N, Musters RJP, Deelman LE, Kamp O. Ultrasound and microbubble-targeted delivery of macro- molecules is regulated by induction of endocytosis and pore forma- tion. Circ Res 2009;104:679 –687.
Netz RR, Schick M. Pore formation and rupture in fluid bilayers. Phys Rev E 1996;53:3875 –3885.
Nieuwoudt M, Engelbrecht GHC, Sentle L, Auer R, Kahn D, van der Merwe S. Non-toxicity of IV injected perfluorocarbon oxygen carrier in an animal model of liver regeneration following surgical injury. Artif Cells Blood Substit Biotechnol 2009;37:117 –124.
Prentice P, Cuschierp A, Dholakia K, Prausnitz M, Campbell P. Membrane disruption by optically controlled microbubble cavita- tion. Nature Physics 2005;1:107 –110.
Rahim AA, Taylor SL, Bush NL, ter Haar GR, Bamber JC, Porter CD. Spatial and acoustic pressure dependence of microbubble-mediated gene delivery targeted using focused ultrasound. J Gene Med 2006; 8:1347 –1357.
Rapoport N, Nam KH, Gupta R, Gao ZG, Mohan P, Payne A, Todd N, Liu X, Kim T, Shea J, Scaife C, Parker DL, Jeong EK, Kennedy AM. Ultrasound-mediated tumor imaging and nanother- apy using drug loaded, block copolymer stabilized perfluorocarbon nanoemulsions. J Control Release 2011;153:4–15.
Rapoport NY, Christensen DA, Fain HD, Barrows L, Gao Z. Ultrasound- triggered drug targeting of tumors in vitro and in vivo. Ultrasonics 2004;42:943 –950.
Schroeder A, Avnir Y, Weisman S, Najajreh Y, Gabizon A, Talmon Y, Kost J, Barenholz Y. Controlling liposomal drug release with low frequency ultrasound: Mechanism and feasibility. Langmuir 2007; 23:4019–4025.
Schroeder A, Kost J, Barenholz Y. Ultrasound, liposomes, and drug delivery: Principles for using ultrasound to control the release of drugs from liposomes. Chem Phys Lipids 2009;162:1–16.
Seymour LW. Passive tumor targeting of soluble macromolecules and drug conjugates. Crit Rev Ther Drug 1992;9:135 –187.
Sheeran PS, Luois S, Dayton PA, Matsunaga TO. Formulation and acoustic studies of a new phase-shift agent for diagnostic and thera- peutic ultrasound. Langmuir 2011;27:10412–10420.
Singh R, Husseini GA, Pitt WG. Phase transitions of nanoemulsions using ultrasound: Experimental observations. Ultrason Sonochem 2012;19:1120 –1125.
Staples BJ, Roeder BL, Husseini GA, Badamjav O, Schaalje GB, Pitt WG. Role of frequency and mechanical index in ultrasonic- enhanced chemotherapy in rats. Cancer Chemother Pharmacol 2009;64:593 –600.
Unger EC, McCreery TP, Sweitzer RH, Caldwell VE, Wu YQ. Acous- tically active lipospheres containing paclitaxel: A new therapeutic ultrasound contrast agent. Invest Radiol 1998;33:886 –892.