*4.3. Lipid Phase Distribution and Acoustical Behavior*

The homogeneous lipid phase distribution in indirect DSPC-cholesterol microbubbles found by 4Pi confocal microscopy was confirmed with confocal microscopy of the microbubbles also analyzed acoustically. Besides the homogeneous lipid phase distribution, indirect DSPC-cholesterol microbubbles had buckles in their coating before insonification

more frequently and more extensively than microbubbles without cholesterol, demonstrated by 4Pi confocal microscopy as well. The DSPC-cholesterol microbubbles insonified at 150 kPa had more buckles than those insonified at 50 kPa, underlining the heterogeneity between different indirect DSPC-cholesterol microbubbles from the same batch. The variable buckle incidence may be explained by the low stability of the DSPC-cholesterol coating. Due to the low amount of LC phase area in their coating, indirect DSPC-cholesterol microbubbles are expected to dissolve at a faster rate than those without cholesterol [41]. Different collapse and shedding mechanisms, such as budding, folding, and buckling, have been proposed to explain how the phospholipid monolayer around the gas core responds to the spontaneous dissolution of the microbubble [42].

Next, the correlation between maximum oscillation amplitude and change in the microbubble coating was investigated. After combining the data from microbubbles insonified at 50 and 150 kPa, the oscillation amplitude of microbubbles that experienced change due to ultrasound insonification was found to be significantly higher than that of unchanged microbubbles for all microbubble types, with a threshold amplitude of approximately 20%. Other studies investigating the lipid coating behavior in microbubbles during ultrasound insonification found comparable results, namely a threshold oscillation amplitude of 30% [43,44]. The difference in threshold amplitude may be explained by the microbubble formulation as microbubbles in other studies were coated with DPPC [43] or DSPC [44] and DSPE-PEG2000, without PEG40-stearate. Another explanation could be a difference in the production method, as the microbubbles for the present study were all made by probe sonication, in contrast to the vial shaker method used for previous studies.

This study is the first to record both the lipid phase distribution and acoustic response in single microbubbles with the combined confocal microscope and Brandaris 128 camera system. Whereas no correlation could be confirmed between the oscillation amplitude and the amount of buckles present before insonification, the LC domain size did correlate with the oscillation amplitude. Microbubbles with small-sized LC domains had higher oscillation amplitudes, which is in accordance with previous reports on phospholipid-coated microbubbles with different sized LC domains, where those with smaller LC domains had a lower resistance to deformation [45]. By contrast, another study found no significant differences in the behavior or stability of microbubbles during ultrasound insonification when they related the lateral molecular packing to the acoustic behavior of microbubbles in different formulations, even though the method of production did affect the lipid packing significantly [46]. The reason for this could be the quantification of the lateral molecular packing, as this was done by calculating the generalized polarization value for single microbubbles, which does not account for lipid phase separation or microstructures in the coating. Before the DSPC-based microbubbles with and without cholesterol studied here can be used for in vivo applications, several differences between the in vitro setting and the in vivo situation need to be considered. Besides the temperature, these differences also include the blood flow, blood viscosity, and soft boundaries affecting the microbubbles' acoustic behavior. Furthermore, the targeting strategy must be adapted to avoid an immune response to streptavidin, a foreign protein [47].
