1.

Introduction

Despite significant advances in the development of molecular therapeutics, effective delivery of these compounds to tissue targets can present a significant challenge. Delivery strategies based on liposomes, nanoparticles, and polymeric complexes have emerged as promising approaches that offer the potential for improved circulation time, targeted delivery, and controlled release.^{1, 2} Effective development of micro-and nanocarrier systems with triggered release would reduce off-target effects and potential toxicities leading to more effective therapies. Liposomes have gained acceptance as a carrier for chemotherapeutics for cancer therapy, but their effective use has suffered from rapid clearance from the circulation via cells of the reticulo-endothelial system (RES), including Kupffer cells in the liver and fixed macrophages in the spleen.³ Liposomes containing polyethylene glycol (PEG; i.e., pegylated STEALTH liposomes), demonstrate reduced uptake by the RES and extended circulation time in vivo.^{1, 4} An example of such a liposomal formulation that is approved for medical use is pegylated liposomal doxorubicin (doxorubicin HCL liposome; Doxil or Caelyx). This structure allows for a pharmacokinetic profile characterized by an extended circulation time and a reduced volume of distribution, which then promotes the uptake by tumors. The associated unique reduction of the volume of distribution of doxorubicin is due to the stable liposomal encapsulation, which allows for an improved safety profile.

Pegylated strategies have, to a large extent, addressed the challenges of limited circulation times, leading to a need for controlled and/or targeted release. Strategies for directed delivery of liposomal contents include the use of targeting ligands or antibodies on the surface or the use of sheddable PEG coatings.⁵ Although PEG prevents uptake of the liposomes by the RES, it also prevents uptake by target cells. This can be overcome by the use of ligands that mediate cell-surface attachment and promote endocytosis, or by removal of PEG at the target site.⁵ Liposome formulations have been reported that release drug contents at temperatures slightly above the normal $37°\mathrm{C}$ body temperature $(42\text{to}45°\mathrm{C})$ .^{6, 7, 8, 9, 10, 11, 12, 13} Directed heating can induce phase changes in the liposomes for targeted release, offering control over time and location of drug delivery.^{6,7,9,10,12,13,14} Potential noninvasive methods for heating the liposomes include laser irradiation, microwave or radio frequency energy, or phased array ultrasound.^{6,10,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29} Here, we have investigated laser-mediated tissue heating and determined that lipid-based carriers released contents at slightly elevated, clinically attainable temperatures $(39\text{to}42°\mathrm{C})$ . Laser activation may enable tissue-specific, controlled release.

The temperature-sensitive liposomes that release drug contents in a clinically obtainable mild hyperthermia range $(39\text{to}42°\mathrm{C})$ have been previously described.^{7, 9, 10} By incorporating water-soluble lysolipid, monopalmitoylphosphatidylcholine (MPPC; $10\mathrm{mol}\%$ ) into dipalmitoylphophatidylcholine (DPPC) liposomes, the phase transition temperature of the liposomes is slightly decreased from $41.9\text{to}\sim 40.5°\mathrm{C}$ (Ref. 10). It has been shown that heating liposomes that have accumulated at tumor sites released the drug contents predominantly in the tumor.¹⁰ Studies of the liposomes for treating human tumor xenograft in mice, in combination with local hyperthermia $(42°\mathrm{C})$ , have shown excellent local controlled release.⁷

An ideal temperature-sensitive liposome vesicle would be one that maintains stability at $37°\mathrm{C}$ during distribution throughout the body. This can be accomplished by the addition of distearoyl-phosphatidylcholine-poly(ethylene glycol-2000) [DSPE-PEG(2000)]. This would be followed by rapid, triggered release via thermal activation at the appropriate temperature $(45°\mathrm{C})$ at the targeted tissue site. Liposomes comprised of DPPC:MPPC:DSPE-PEG(2000) with an $86:10:4\mathrm{mol}\%$ exhibited stability at body temperature and released the encapsulated drug rapidly, on the order of $>80\%$ within tens of seconds after thermal activation.^{7, 8, 9}

The encapsulation and leakage of the fluorescent dye carboxyfluorescein (6-CF and 5,6-CF) from liposomes is one of the commonly used methods to investigate stability and release.^{3, 10, 12, 16, 17, 18, 19, 20, 28, 30, 31, 32} To provide such similar readout or a signal switch in living cells and tissues, we used the luciferin–luciferase reaction with luciferin as a model small molecule. Since this reaction is limited to the cytoplasm, the bioluminescent signal indicates release of luciferin from the liposomes and cellular uptake.

Transgenic reporter mice, which were engineered to ubiquitously and constitutively express firefly luciferase in all tissues,^{33, 34} were used to noninvasively reveal spatiotemporal biodistribution of luciferin using in vivo bioluminescent imaging (BLI). BLI has been shown to be extremely sensitive, and the use of the luciferin–luciferase reaction provides a signal switch for intracellular delivery. Signal to noise is relatively high with BLI due to extremely low background, which offers advantages over the use of fluorescent reporters.^{35, 36} In this manner, BLI can be used for monitoring the effectiveness of drug treatment and provides a broad dynamic range for quantification with acceptable resolution for tissue localization.³⁷

Use of luciferin as a model drug allows for the quantification of its release from the liposome after laser activation. The photons emitted from the transgenic mouse provide an improved method of analysis based on BLI. Studies reported by Kim have demonstrated thermal release of luciferin after local delivery.³⁸ However, the purpose for thermal activation is to enable local release after systemic delivery. The use of DSPE-PEG(2000)-PDP in this study likely limited this liposome preparation to local delivery, since this lipid has a sulfhydryl group for the purpose of functionalization of the lipid surface.³⁸ By not utilizing this sulfhydryl for functionalization, or otherwise reducing it, this group would remain available for forming disulfide linkages between liposomes, cells, and the endothelial surface, thus limiting its use to local injection. Since the purpose of thermal activation is to locally release systemically delivered drug, our objective was to use a liposome formulation that would allow systemic delivery and local release.³⁹ By using a transgenic mouse model and luciferin encapsulated in a liposome vesicle, we were able to assess the extent and location of release of liposomal contents in real time at a desired location after intravenous delivery. The capability for real-time in vivo assessment of delivery of the liposome payload provides an effective tool for optimizing liposome composition and assessing effective delivery.

2.

Materials and Methods

3.

Results and Discussion

In this study, we used D-luciferin as a model drug. This compound, when used with transgenic reporter mice, allows for imaging the location of targeted delivery of model drug and can be used to assess laser-mediated thermal release from liposomes. The luciferin-containing liposomes in the tissues were heated using a Nd: YLF $\left(527\mathrm{nm}\right)$ laser source, and the in vivo spatiotemporal biodistribution of luciferin was imaged using BLI. In addition, the onset of release for the entrapped contents (in this case D-luciferin) occurs more rapidly than for the pure DPPC lipid at temperatures just above $39°\mathrm{C}$ . At high concentrations (greater than $30\mathrm{mM}$ ), the traditionally used carboxyfluorescein dye is self-quenched and emits a very low fluorescent signal. When released from the liposomes, the fluorescence signal can be measured with time due to dilution into a larger, external volume. Although useful for assessing liposome stability, this approach cannot be used to indicate when the liposomal contents have been released and have crossed the cellular membrane into the cytoplasm of the target cell. As far as we know, this is the first report that shows the controlled release of D-luciferin from a liposome at the intended target after intravenous injection. As the L2G85 transgenic mouse model constitutively expresses luciferase, we can effectively quantify the release of the luciferin at the heated site, and any off-target tissues, using BLI measurements. In this study, we used the MPPC:DPPC:DSPE-PEG(2000) liposome which is reported^{7, 10, 11, 12, 13} to have both a reasonable in vivo half-life due to the protection from the RES, and good thermal release kinetics at $\sim 42°\mathrm{C}$ . In addition, we were able to show that a Nd: YLF laser operating at $527\mathrm{nm}$ in wavelength was able to release the IV injected small molecule, luciferin, within the target tissue by transiently heating the region to $45°\mathrm{C}$ at the skin surface.

Luciferin release from the liposomes was tested at temperatures of 37 and $45°\mathrm{C}$ in both PBS and serum using bioluminescence measurements with the luminometer as described in the methods. The liposomes combined with PBS were heated from $37\text{to}45°\mathrm{C}$ for $5\min$ on the day that they were made. The release of luciferin is reported as fold increase in bioluminescence, which was calculated by dividing the average bioluminescence at $45°\mathrm{C}$ by the average bioluminescence at $37°\mathrm{C}$ . A 3.5-fold increase of bioluminescence was seen for the liposomes in PBS, and a 2.5-fold increase was seen for the liposomes in serum. The results are shown as an average in photons/sec for day zero between PBS and serum at the two different temperatures in Fig. 1a .

Fig. 1

Liposome stability in vitro. (a) The average photons/second $(N=3)$ of luminescence from the designed $86:10:4\text{molecular}$ weight percent of the MPPC:DPPC:DSPE-PEG(2000) liposome containing D-luciferin is shown ex vivo for analysis. The data is shown for a 1:10 dilution of the liposome by adding $10\mu \mathrm{l}$ of liposome to $90\mu \mathrm{l}$ of PBS or serum. The liposome was thermally heated to either 37 or $45°\mathrm{C}$ for comparison. (b) The average fold change $(N=3)$ of bioluminescence from the D-luciferin containing liposome from a 1:10 dilution in PBS between 37 and $45°\mathrm{C}$ at 0, 1, 13, and $105\text{days}$ after formation to determine the stability of the liposome over time. (c) The average fold increase $(N=4)$ for bioluminescence after intravenous injection of $100\mu \mathrm{l}$ of D-luciferin containing liposomes following $2\text{-}\min$ water bath heating at $45°\mathrm{C}$ on the heated right paw and the unheated left paw. The data was normalized to the bioluminescence of the whole body. The thermal release and imaging was performed $40\min$ after intravenous injection. Imaging was performed with $1\text{-}\min$ integration. The error bars show ± the standard deviation of the normalized signal.

The next experiment compared the release of luciferase from the liposomes after storage. Bioluminescent signals were measured at $37°\mathrm{C}$ and then heat treated $(45°\mathrm{C})$ in PBS on days 0, 1, 13, and 105 to determine the long-term stability of the liposome prior to injection. Results are reported as fold change in luciferase activity measured at $37°\mathrm{C}$ and $42°\mathrm{C}$ . A 3.7-fold increase in signal was noted at day 1, and a 2.7-fold increase was seen at day 13 following formulation. By day 105, a 1.7-fold increase of bioluminescence for liposomes/PBS was observed at the two temperatures (Fig. 1). In addition to these long-term results, Needham have shown the ability of PEG containing liposomes to have a greater stability due to the hydrophilic properties of PEG, which should translate to long term stability in vivo.^{7, 8, 40} While luciferin is maintained inside these liposomes at room temperature $(25°\mathrm{C})$ and body temperature $(37°\mathrm{C})$ , increased temperatures release the liposomal contents, and once luciferin is released, it permeabilizes through nearby cell membranes due to the amphipathic structure of this molecule.^{41, 42, 43, 44}

For in vivo studies, $100\mu \mathrm{l}$ of liposome solution was injected intravenously into transgenic reporter mice. These mice express luciferase constitutively throughout the body and provide an excellent platform for investigating the biodistribution of luciferin and directed delivery. The liposomal release was first tested $40\min$ post–IV injection by placing the right paw of the transgenic mouse into a $45°\mathrm{C}$ water bath for $2\min$ . Some background bioluminescence was observed throughout the mouse after IV injection as a result of luciferin that had either leaked out or was not encapsulated. The thermal activation studies were performed $40\min$ after the IV injection of the liposomes, which was determined to be a time when the background signals were low. Each mouse was imaged in the IVIS200 (Caliper/Xenogen) immediately after heating. Analyses of the in vivo images were performed using LivingImage 2.5 (Caliper). An $n=4$ was used for comparison of mice. The results showed a 1.5-fold increase in signal on the right paw after heating and a 0.6-fold increase in the left paw when normalized to the whole body signal [Fig. 1c].

A Nd:YLF laser was used at $527\mathrm{nm}$ to heat the right foot of mice $90\min$ after injection of the luciferin-liposomes. The $50\text{-}\min$ difference between this experiment and the water bath experiment was due to the location of the laser relative to the imaging system. The laser was some distance from the imaging system and required transfer of the mice and the $50\text{-}\min$ delay. Given that the $40\text{-}\min$ background data indicated that the background had reached a plateau, the delay in imaging likely did not affect background signals but may have reduced signals from luciferin released from liposomes. The right hind feet of the mice were heated to $45°\mathrm{C}$ for $2\min$ as determined by a FLIR camera [FLIR System, Boston, MA, Fig. 2a ]. The first minute of laser irradiation raised the skin temperature from $35\text{to}45°\mathrm{C}$ , while the temperature was maintained at $45°\mathrm{C}\pm 1°\mathrm{C}$ for an additional minute. The laser heat treatment was appropriately guided with the use of a shutter (Vincent Associates) and the FLIR camera. Figure 2b shows the average temperature of the right paw in the $x$ and $y$ direction at the end of the first minute of laser heating. The background signal had dissipated by $90\min$ post-injection [Figure 3a ; Pre-Heat]; therefore, laser heating and imaging was accomplished at $90\min$ post-injection for each of the four mice used for this experiment.

Fig. 2

Thermal analysis of liposome laser release. (a) Thermal imaging of the right paw during Nd:YLF laser irradiation at $650\mathrm{mW}$ , $240\mathrm{Hz}$ , with a $1\text{-}\mathrm{cm}$ elliptical laser spot is shown from the FLIR mid-infrared thermal imaging system. LI01 shows the Y line, and LI02 shows the X line for analysis of the temperature across the elliptical laser spot on the mouse paw. The laser energy raised the temperature of the paw to $45°\mathrm{C}$ within $1\min$ and was maintained at $45°\pm 1°\mathrm{C}$ with a shutter maintaining the temperature by continually blocking the laser energy during one additional minute. The scale bar shows the temperature relating to each color throughout the image. (b) The temperature in degrees Celsius between the X and Y lines at the end of the first minute after the $45°\mathrm{C}$ temperature was achieved on the right paw.

Fig. 3

In vivo imaging of the release of liposome contents. (a) Image of one of the laser-irradiated mice. This imaging was done with an IVIS50 imaging system (Caliper/Xenogen, Inc.) with a $1\text{-}\min$ integration time. The imaging was done $1\min$ prior to laser heating, immediately after laser heating, and $3\min$ after laser heating. The image shows a clear difference between the bioluminescence signal seen in the laser heated right paw and the unheated left paw. (b) The average fold increase $(N=4)$ in bioluminescence signal after intravenous injection of $100\mu \mathrm{l}$ of D-luciferin containing liposomes following $2\text{-}\min$ laser heating to $45°\mathrm{C}$ on the heated right paw and the unheated left paw. The data was normalized to the bioluminescence of the whole body. The thermal release and imaging was performed $90\min$ after intravenous injection. Imaging was performed with $1\text{-}\min$ integration. The error bars show ± the standard deviation of the normalized signal. (c) A comparison between the average fold induction between the water bath and the Nd:YLF laser system is shown. The data was normalized to the whole body region of interest and the left paw in each mouse imaged. The final result shows a 3-fold increase in bioluminescence signal related to the laser heating compared to a 2.5-fold increase related to the water bath heating.

An IVIS50 imaging system (Caliper LifeSciences, New York) was used to image the mice. Differences in luciferase activity have been noted between the temperatures of 25 and $37°\mathrm{C}$ (Ref. 45). Over this $12\text{-}\text{degree}$ temperature range, there is only a 0.2-fold increase in activity. Since the FLIR camera indicated that the tissue return to normal temperature by the time the animals are imaged, and the modest differences in activity that may be possible at elevated temperatures, it is evident that increases in activity after heating are due to thermal release. In this study, the normal, $37°\mathrm{C}$ , temperature was achieved within a few seconds after heating and well before the imaging began.^{46, 47} The image of one laser-heated mouse is shown in Fig. 3a. The mouse is shown prior to laser heating at $90\min$ post-injection of liposomes (IV) and immediately after laser heating, as well as at $3\min$ after laser heating. A 2.6-fold increase in bioluminescence in the heated right paw was seen compared to a 0.8-fold increase in the left paw [Fig. 3b]. Figure 3c shows a normalization between the heating of the liposome for $2\min$ with the water bath and the Nd:YLF laser. This data was normalized to the whole body region of interest and the left paw region of interest for each mouse prior to averaging with an $N=4$ for each experiment. The laser heating showed a normalized 3-fold increase in bioluminescence compared to a 2.5-fold increase due to water bath heating.

4.

Conclusions

We have tested stable, temperature-sensitive liposomes that release their drug contents upon heating for the ability to administer the liposomes systemically and activate them locally in a target tissue. The technique was characterized and optimized by employing luciferin as a method for in vitro and in vivo tracking and analysis. The tested formulations have shown encouraging results with regards to stability and release profile upon heating in vitro in PBS and serum. We were able to thermally release the liposome contents with both a water bath and after heating with a Nd:YLF $527\mathrm{nm}$ laser. The laser showed a 3-fold increase in bioluminescence compared to a 2.5-fold increase following heating in a water bath. We have shown that using D-luciferin as a model molecule and employing the transgenic reporter mouse as an indicator of drug release by BLI is a useful strategy for developing and optimizing a range of liposomes and other drug delivery methods.^{38, 48, 49} Such a system will allow us to develop better drug delivery tools for improved control of drug release in target tissues. While every drug has unique properties, it is possible to use this system for analyzing delivery tools that may normalize these differences.