Date Published: July 7, 2017
Publisher: Public Library of Science
Author(s): Klazina Kooiman, Tom van Rooij, Bin Qin, Frits Mastik, Hendrik J. Vos, Michel Versluis, Alexander L. Klibanov, Nico de Jong, Flordeliza S. Villanueva, Xucai Chen, Juan TU.
Acoustic behavior of lipid-coated microbubbles has been widely studied, which has led to several numerical microbubble dynamics models that incorporate lipid coating behavior, such as buckling and rupture. In this study we investigated the relationship between microbubble acoustic and lipid coating behavior on a nanosecond scale by using fluorescently labeled lipids. It is hypothesized that a local increased concentration of lipids, appearing as a focal area of increased fluorescence intensity (hot spot) in the fluorescence image, is related to buckling and folding of the lipid layer thereby highly influencing the microbubble acoustic behavior. To test this hypothesis, the lipid microbubble coating was fluorescently labeled. The vibration of the microbubble (n = 177; 2.3–10.3 μm in diameter) upon insonification at an ultrasound frequency of 0.5 or 1 MHz at 25 or 50 kPa acoustic pressure was recorded with the UPMC Cam, an ultra-high-speed fluorescence camera, operated at ~4–5 million frames per second. During short tone-burst excitation, hot spots on the microbubble coating occurred at relative vibration amplitudes > 0.3 irrespective of frequency and acoustic pressure. Around resonance, the majority of the microbubbles formed hot spots. When the microbubble also deflated acoustically, hot spot formation was likely irreversible. Although compression-only behavior (defined as substantially more microbubble compression than expansion) and subharmonic responses were observed in those microbubbles that formed hot spots, both phenomena were also found in microbubbles that did not form hot spots during insonification. In conclusion, this study reveals hot spot formation of the lipid monolayer in the microbubble’s compression phase. However, our experimental results show that there is no direct relationship between hot spot formation of the lipid coating and microbubble acoustic behaviors such as compression-only and the generation of a subharmonic response. Hence, our hypothesis that hot spots are related to acoustic buckling could not be verified.
Ultrasound contrast agents (UCAs) consist of coated gas microbubbles (1–10 μm in diameter) dispersed in an aqueous suspension. These blood pool agents aid in the diagnosis of for example liver  and kidney lesions  and in left ventricular visualization . In the blood pool, uncoated microbubbles would dissolve in less than 0.3 s  which is too short a lifetime for diagnostic imaging; a coating is therefore essential for increased stability and thus longevity of the microbubbles. The coating reduces the surface tension and the corresponding capillary pressure that drives the gas out of the microbubble core into the surrounding fluid. In addition, it forms a barrier that reduces gas diffusion [5–7].
The number weighted mean microbubble diameter as determined from the Coulter Counter measurements was 3.7 μm with a standard deviation of diameter of 2.6 μm. In total, 137 randomly selected microbubbles that met the inclusion and exclusion criteria were studied optically, where the smallest microbubble had a diameter of 2.3 μm and the largest microbubble was 10.3 μm in diameter. About half of the microbubbles (51.8%; n = 71) showed inhomogeneities in the fluorescent coating before insonification. These were defined as focal areas of increased fluorescence; hereafter referred to as “hot spots”. Interestingly, the occurrence of hot spots was higher in smaller microbubbles (D0 < 6 μm) as shown in Fig 1A. On average 2.2 hot spots per microbubble coating were observed, with a range of one to five as shown in Fig 1B. Typical examples of these hot spots before ultrasound application are given in Fig 2B2, 2B4, and 2B6 (leftmost column) for three different microbubbles. The microbubble in Fig 2B2 had four hot spots, the one in Fig 2B4 three, and the microbubble in Fig 2B6 had two hot spots, all indicated by arrow heads. To the best of our knowledge, this is the first study that investigates on a nanoseconds time scale the dynamic lipid motion in the microbubble coating during insonification. We observed three different types of behavior of the fluorescently labeled lipid coating: (a) no change in fluorescence; (b) reversible hot spot formation during insonification (only in compression phase); (c) irreversible hot spot formation during insonification (in compression and expansion phase that persisted after ultrasound was turned off). Hot spots were first formed in the compression phase when the relative vibration of the microbubble was > 0.3, irrespective of the insonification frequency (0.5 or 1 MHz) and P_ (25 or 50 kPa).
Using ultra-high-speed fluorescence recordings, we observed the formation of focal areas of increased fluorescence or hot spots, on the lipid monolayer microbubble coating at relative vibrations > 0.3 at a frequency of 0.5 and 1 MHz at a P_ of 25 and 50 kPa. Around resonance, the majority of the microbubbles formed hot spots. Formation of hot spots was always observed in the compression phase and in 68% of the cases they also persisted in the expansion phase and after the ultrasound was turned off. If the microbubble also acoustically deflated, hot spot formation was likely irreversible. While we have observed that acoustic vibration leads to the formation of hot spots, we did not find a correlation of hot spot formation with nonlinear acoustic behavior of the microbubble. Therefore, we could not verify the previous hypothesis that monolayer collapse by buckling or folding of the lipid coating of the microbubble on a molecular scale leads to nonlinear acoustic behavior of the microbubble.