Ultrasonics 55 (2015) 1–5
Contents lists available at ScienceDirect
Ultrasonics journal homepage: www.elsevier.com/locate/ultras
Temporal effect of inertial cavitation with and without microbubbles on surface deformation of agarose S gel in the presence of 1-MHz focused ultrasound Y. Tomita a,⇑, T. Matsuura a, T. Kodama b a b
Faculty of Education, Hokkaido University of Education, Hakodate, Hokkaido 040-8567, Japan Graduate School of Biomedical Engineering, Tohoku University, 4-1 Seiryo, Aoba-ku, Sendai 980-8575, Japan
a r t i c l e
i n f o
Article history: Received 8 May 2014 Received in revised form 25 July 2014 Accepted 27 July 2014 Available online 4 August 2014 Keywords: Sonoporation Inertial cavitation Microbubbles Ultrasound Surface roughness
a b s t r a c t Sonoporation has the potential to deliver extraneous molecules into a target tissue non-invasively. There have been numerous investigations of cell membrane permeabilization induced by microbubbles, but very few studies have been carried out to investigate sonoporation by inertial cavitation, especially from a temporal perspective. In the present paper, we show the temporal variations in nano/micro-pit formations following the collapse of inertial cavitation bubbles, with and without SonazoidÒ microbubbles. Using agarose S gel as a target material, erosion experiments were conducted in the presence of 1-MHz focused ultrasound applied for various exposure times, Tex (0.002–60 s). Conventional microscopy was used to measure temporal variations in micrometer-scale pit numbers, and atomic force microscopy utilized to detect surface roughness on a nanometer scale. The results demonstrated that nanometerscale erosion was predominantly caused by SonazoidÒ microbubbles and C4F10 gas bubbles for 0.002 s < Tex < 1 s, while the number of micrometer-scale pits, caused mainly by inertial cavitation bubbles such as C4F10 gas bubbles and vapor bubbles, increased exponentially with increasing Tex in the range 0.1 s < Tex < 10 s. The results of the present study suggest that cavitation-induced sonoporation can produce various pore sizes in membranes, enabling the delivery of external molecules of differing sizes into cells or tissues. Ó 2014 Elsevier B.V. All rights reserved.
1. Introduction Cell membrane sonoporation is enhanced by the addition of shell-encapsulated microbubbles such as ultrasound contrast agents (UCAs) [1,2]. Numerous studies have investigated the interactions of inertial cavitation bubbles with cells in the presence of ultrasound (US), in order to elucidate the mechanisms of sonoporation and related biological effects [3,4]. Shear stress arises from microstreaming induced by the motion of a microbubble, and is recognized as the dominant cause of membrane permeabilization and sonoporation that results from the interaction of low-intensity US with microbubbles [5,6]. However, microbubbles can also behave as inertial cavitation bubbles when an interaction between UCAs and a boundary occurs in the presence of either moderate- or high-intensity US [6,7]. Forbes et al.  found a correlation between ⇑ Corresponding author. Address: Faculty of Education, Hokkaido University of Education, Hakodate, 1-2 Hachiman-cho, Hakodate 040-8567, Japan. Tel./fax: +81 (0)138 44 4307. E-mail addresses: [email protected]
(Y. Tomita), matsuura.toshi [email protected]
(T. Matsuura), [email protected]
(T. Kodama). http://dx.doi.org/10.1016/j.ultras.2014.07.017 0041-624X/Ó 2014 Elsevier B.V. All rights reserved.
the number of ruptured Optison microbubbles and increased sonoporation in Chinese hamster ovary (CHO) cells. Prentice et al.  used optically-controlled microbubble cavitation to observe UCA microjets and the resultant pits, and concluded that sonoporation at high-intensity US may arise from a synergistic interplay of several different processes. Caskey et al.  found that individual microbubbles migrated towards the microvessel wall and were sometimes accompanied by associated microjets. However, Chen et al.  reported that US-activated microbubbles migrated away from the nearest vessel wall. These results suggest that different stresses can occur on the surface of a cell or tissue depending on the state of the acoustic ﬁeld, the bubble characteristics and the dynamic nature of the nearby boundary. Chomas et al.  have observed sonic cracking of a contrast agent microbubble. The gas that ﬂows to the exterior from a disrupted microbubble will either become free to grow as a bubble or dissolve in the liquid [6,12]. Shell fragments in the liquid can act as seeds for cavitation, i.e. cavitation nuclei. Thus, in the presence of an acoustic ﬁeld generated by moderateor high-intensity US, there are many opportunities for gas bubbles, created by shell rupture, to grow into larger bubbles through
Y. Tomita et al. / Ultrasonics 55 (2015) 1–5
rectiﬁed diffusion. Although there have been numerous investigations of microbubble-induced cell membrane permeabilization, very few have studied sonoporation induced by cavitation bubbles that are generated after the destruction of microbubbles. Clear evidence of sonoporation has been obtained in experiments showing successful delivery of ﬂuorescent molecules into cells. However, small pores in cells induced by UCA microjets can disappear within a few seconds due to membrane sealing . Thus, accurately elucidating the mechanisms of sonoporation is challenging, especially from a temporal perspective when US exposure times of up to several tens of seconds are used. SonazoidÒ microbubbles are commercially-available encapsulated UCAs used mainly for hepatic blood ﬂow imaging . In the present study, erosion experiments were conducted in which 1-MHz focused ultrasound was applied for various exposure times, Tex (0.002–60 s), in the presence or absence of SonazoidÒ microbubbles, and the temporal variations in nano/micro-pit formation following the collapse of inertial cavitation bubbles were determined. Agarose S gel was used as the target material instead of living cells or tissues in order to register the evidence of stresses as accurately as possible. Atomic force microscopy (AFM) was utilized to detect surface roughness on a nanometer scale, and conventional microscopy used to analyze the temporal variation in micrometer-scale pit numbers and sizes.
2. Materials and methods Fig. 1 shows a schematic illustration of a typical test section. It consisted of the test material and a 1% agar gel vessel that contained a cylindrical space of 10 mm diameter and 5 mm height, fabricated by molding the upper portion of the gel; the cylindrical space contained a 5% Sonazoid suspension (i.e. volumetric concentration, Cs = 5%) at 25 °C. SonazoidÒ for injection is a commercially available, second-generation, encapsulated UCA used mainly for hepatic blood ﬂow imaging (Daiichi Sankyo Co. Ltd., Tokyo, Japan). SonazoidÒ microbubbles have an average diameter of 2.1 ± 0.7 lm, and are comprised of a 4-nm thick H-EPSNa shell that conﬁnes C4F10 gas, i.e. perﬂubutane. Assuming that each microbubbles was spherical, the initial number of SonazoidÒ microbubbles per unit volume was calculated to be N5% = 6.3 104 particles/mm3 for Cs = 5%. The void fraction, a, was calculated to be 4.0 10 4. We used 5% agarose S gel (AS-gel) as a ‘model’ of the boundary of an internal organ consisting of many cells. Although agarose S gel is a mesh of macromolecules, and thus differs from a cell membrane comprised of a very thin phospholipid bilayer, it is an appropriate target material for investigating, as accurately as possible, the temporal variations in surface deformation that occur in response to various stresses. If cells or tissues were used as the target material, membrane sealing might occur within a few seconds of irradiation with US, leading to an incorrect
55 µm 0.28 0.21 mm
Fig. 1. Schematic illustration of the test section. MS area (=0.060 mm2) and AFM area (=25 lm2) represent the observation areas viewed by conventional and atomic force microscopy, respectively.
evaluation of pore numbers. Agarose S is a polysaccharide with a polymeric network structure, and has a melting point of 90 °C and a gel–sol transition temperature of 37–39 °C (1.5% AS-gel). The root mean square roughness, Rrms, of the AS-gel virgin surface was 8 ± 2 nm. Focused US at a frequency of 1-MHz was produced continuously for various exposure times, Tex (0.002–60 s), using a multipurpose synthesizer with a bipolar ampliﬁer, and a concave transducer with a chord radius, a, of 20 mm and a geometric focal length, R, of 80 mm . A substantial interaction time, TR, between focused US and SonazoidÒ microbubbles was deﬁned as Tex DR, where DR is the duty ratio. Burst waves with a duty ratio of 99.99% were mostly used; in this paper, they have been termed continuous waves (CW). The maximum pressure, pmax, at the US focus was deﬁned as the half-value of the full amplitude of the maximum pressure signal among the leading 20 waves, whose absolute value was equal to the largest negative pressure, p_. The full width half maximum (FWHM) of pmax, given by 2kR/(pa), was calculated to be 3.8 mm for a wavelength, k, of 1.5 mm. For erosion experiments, pmax was ﬁxed at 0.75 MPa (±15%). The initial acoustic intensity was calculated as 18.8 W/cm2. The damaged surface was observed using a conventional microscope with an observation area of 0.060 mm2, while a scanning probe microscope (Shimadzu SPM-9500J3) was utilized to obtain AFM images on a nanometer scale. A silicon probe with an edge radius of 10 nm, mounted on the head of a cantilever operating in dynamic mode at a frequency of 300 kHz, was used to identify the surface. The AFM scanning area was 25 lm2 with a resolution of 512 512 pixels. To obtain information about bubble motion, high-speed photography was used at a maximum rate of 300,000 frames/s (Photron Ltd. FASTCAM SA5). This temporal resolution was not sufﬁcient to observe the radial motion of a SonazoidÒ microbubble with a diameter of 2.1 lm, but was partially capable of capturing the translational motion of individual microbubbles, except in cases where the oscillatory translational motion of the microbubble was more rapid. The temperature variation in the ASgel, at a point 2 mm distant from the surface where the US was focused, was examined preliminarily with a thermocouple made of copper-constantan; each wire was 0.1 mm in diameter, and the soldered tip diameter was about 1 mm.
3. Results and discussion During a brief period following US irradiation, standing waves were formed near the target surface since there was very little ﬂow in the water layer in response to acoustic radiation pressure. Subsequently, SonazoidÒ microbubbles arose very slowly, joining the acoustic streaming induced by the non-linear effect of the ﬂow. However, when water ﬂow became evident near the surface, the liquid pressure in the water ﬂuctuated, resulting in a break in the state of standing waves. Several tens of milliseconds later, microbubble motion became more active to yield coalescence that occurred due to shell rupture during expansion; this mechanism differs from that for shell-free bubbles, where Van der Waal attractive forces are important for coalescence [7,12]. From our overall observations, coalescing microbubbles with a modal diameter of 10.8 lm were born near the gel surface during a 6 ms period, migrated towards the surface at a velocity of 1–4 m/s, and ﬁnally adhered to it. The attached bubble tended to stay in place, oscillating, for a relatively long period of time. Shear stress could occur at the gel surface near a liquid layer that vibrated due to microstreaming of the bubble. Thus, the total number of microbubbles positioned near the gel surface was reduced due to adhesion to the surface, coalescence and destruction of the microbubbles. The remaining number of SonazoidÒ microbubbles, N, decreased with Tex, and fell to <40% of the initial number, N0 (=6.25 104 particles/mm3 for Cs = 5%), at Tex = 0.1 s; N/N0 approached zero at
Y. Tomita et al. / Ultrasonics 55 (2015) 1–5
Tex = 1 s. Fig. 2 shows two trimmed photomicrographs taken for two different exposure times, Tex = 0.002 s (a) and 5 s (b), and a log–log scale diagram (c) indicating the number of pits per unit area, Np (pits/mm2), as a function of Tex; solid circles correspond to the Sonazoid concentration, Cs = 5%, and open circles to Cs = 0% (i.e. the distilled water control). All images were divided into observation areas, 0.060 mm2, in order to obtain an averaged pit number over each area. When distilled water was used, no pitting was observed in the range Tex < 0.1 s; in the range Tex > 0.1 s, pits were produced by inertial vapor cavitation only. On the other hand, when 5% Sonazoid suspensions were used, US irradiation allowed the SonazoidÒ microbubbles to behave as inertial cavitation bubbles. Each bubble grew to around the resonant bubble diameter through a process of bubble coalescence; sometimes bubbles reached a diameter more than several times that of the resonant bubble diameter (9.2 lm) corresponding to the 1-MHz frequency of the applied US , due to non-linear bubble oscillation. Collapse of an attached bubble was involved in the production of a microjet with a diameter about one tenth of the maximum bubble diameter [16,17]. Consequently, various sizes of shallow depressions and circular pits occurred on the AS-gel surface, as shown in Fig. 2 for Tex = 0.002 s (a) and Tex = 5 s (b). Typical pits, of diameters 3 lm and 4 lm, respectively, are arrowed by A in (a) and B in (b). The total number of pits detected in the observation area, 0.060 mm2, of a photomicrograph corresponding to (b) exceeded more than 300, with pit diameters ranging from 1 to 7 lm. Naturally, the number of C4F10 gas bubbles increased with Tex because the SonazoidÒ microbubbles were being destroyed continuously. Fig. 2(c) is a log–log graph of pit numbers per unit area, Np, plotted against Tex; the two solid lines for the range 0.1 s < Tex < 10 s are empirical curves derived from power-law ﬁtting, where Np = aTbex; and a = 57 and b = 0.75 for Cs = 0%, and a = 1258 and b = 0.72 for Cs = 5%. Np reached a maximum at around Tex = 10 s, where pit diameters of about 10 lm were obtained, and decreased for Tex > 10 s owing to thermal effects. Thermal conductivity of AS-gel is smaller than that of water, and there is no heat convection inside it; as a result, heat would have accumulated in the AS-gel as Tex increased, with the temperature temporarily exceeding 60 °C at Tex = 40 s. For longer exposure times, the AS-gel surface became softened. Overall, the observations suggest that inertial C4F10 gas bubbles may play an important role in erosion for Tex > 0.1 s, where about 60% of the SonazoidÒ microbubbles are destroyed . In conjunction with an increase in dissolved C4F10 gas, the inception of vapor cavitation will also be promoted when the liquid temperature increases, due to a rise in the saturated vapor pressure. Although micrometer-scale pits were detected, the AS-gel surface would be expected to be exposed to various ﬂows and impulsive pressures when interactions between bubbles and
focused US occurred near a boundary. Fig. 3 shows three AFM images taken at different exposure times, Tex: (a) 0.5 s, (b) 1 s and (c) 20 s. The cross-sectional curve scanned across the centerline is displayed above each image. In an AFM image, the color brown implies a ﬂat surface, black a depressed surface or pit, and white a projecting surface. Many circular- or oblong-shaped small pits appeared as Tex increased, with pit sizes ranging from several tens of nanometers to 200 nm, as shown in Fig. 3(a) and (b). Small and slender pits with sizes of around 100 nm were also detected, as illustrated by the dashed white circle in Fig. 3(a). These slender depressions presumably were produced by unidirectional shear stress loading due to microstreaming [5,6,18]. For longer exposure times, features of cumulative erosion were revealed, with small pits hidden owing to the subsequent development of larger pits. As shown in Fig. 3(c), the surface became rough and contained large-scale concave and convex impressions. Fig. 4 shows a summary of the erosion characteristics. It is clear that when distilled water was used as a test liquid, the exposure time had little effect on surface roughness. In contrast, when a Sonazoid suspension was used, the erosion characteristics could be classiﬁed into two categories. In the ﬁrst category, corresponding to the range 0.002 s < Tex < 1 s, erosion was likely dominated by impulsive pressures attributable to the collapse of inertial cavitation bubbles such as C4F10 gas bubbles or SonazoidÒ microbubbles. The solid curves representing Rrms and Rv (the maximum valley depth deﬁned as the maximum depth of a depression or pit) resulted from the collapse of inertial SonazoidÒ microbubbles. Each curve had a maximum at a Tex just below 1 s, and rapidly decreased to zero at around 1 s because of complete destruction of the SonazoidÒ microbubbles . Although the maximum value of Rv was less than 70 nm, this is larger than the thickness of many common cell membranes (circa 7–10 nm). It is noteworthy that there are two dashed curves for Rrms and Rv, which may result from the collapse of inertial cavitation bubbles such as C4F10 gas bubbles and/or vapor bubbles. It should also be noted that the pit appearance on the control surface exhibited stochastic features because the control surface area measured by AFM was very small, 25 lm2, and pressure loadings occurred randomly, both temporally and spatially. The second category involved the range Tex > 1 s: the erosion curve was characterized only by the dashed lines shown in Fig. 4(a) and (b). In this region, effective propagation of US waves through the liquid layer resulted in a higher temperature increase at the AS-gel surface. The saturated vapor pressure will rise inexorably with an increase in the temperature of the liquid, resulting in a decrease in the threshold for commencement of cavitation . As a result, the pitting erosion that dominated when Tex > 1 s was probably caused by the collapse of inertial cavitation bubbles such as C4F10 gas bubbles or vapor bubbles. As further time elapsed, erosion was dependent on the cavitation effect coupled with thermal effects. Fig. 4(c) shows the number of pits, Np, per lm2; the dashed line
B 10 μm
Fig. 2. Two trimmed photomicrographs of eroded surfaces for US exposure times, Tex: (a) 0.002 s and (b) 5 s. The scale bars in (a and b) each represent 10 lm. The four arrows in (a) denote shallow depressions. A log–log graph of pit numbers per mm2, Np, against Tex is shown in (c). The solid circles show data for a 5% Sonazoid suspension (Cs = 5%) and the open circles represent data for distilled water (Cs = 0%). The error bars correspond to the standard deviations, and the two solid lines imply empirical curves derived from power-law ﬁtting. The dashed line denotes a critical value of Tex below which no pits were generated for distilled water.
Y. Tomita et al. / Ultrasonics 55 (2015) 1–5
Rrms= 9.2 nm; 5.0×5.0 µm
Rrms= 9.7 nm; 5.0×5.0 µm
Rrms= 29.1 nm; 5.0×5.0 µm
Fig. 3. AFM images and cross-sectional proﬁles scanned at the center line of each image, at three exposure times, Tex: (a) 0.5 s, (b) 1 s and (c) 20 s (Cs = 5%). The cross-sectional curves scanned across the centerline (denoted by the white solid line) are shown above each image. The scale bar denotes 1 lm.
Fig. 4. Nanoscale characteristics of erosion by inertial cavitation as a function of Tex. The solid circles represent data for the 5% Sonazoid suspension, and open circles data for distilled water. (a) Surface roughness expressed as the root mean square, Rrms; (b) maximum depth of the valley, Rv; and (c) pit number per unit area, Np. In (a) and (b), the solid line implies the results from the collapse of SonazoidÒmicrobubbles and/or C4F10 gas bubbles, and the dashed line the results from collapse of either C4F10 gas bubbles or vapor bubbles. Rrms,0 (=8 nm), Rv,0 (=39 nm) and Np,0 (=3.7 pits/lm2) were the initial values for surface roughness, maximum depth of the valley and pit number, respectively.
represents the average number of pre-existing pits (3.7 pits/lm2). More than twice the number of pre-existing pits was detected over a wide range of Tex. For Tex > 1 s, the pit numbers decreased but the surface became rougher. It has therefore been demonstrated unequivocally that the addition of SonazoidÒ microbubbles to a liquid enhances the number of nanometer-scale pits on a gel surface, as compared with a distilled water control for which <5 pits/lm2 were detected at all Tex tested.
4. Conclusion Although there can be no doubt that focused US is an essential factor affecting surface erosion, mixed forces induced by bubble motions, especially shear stress due to microstreaming and jet impact caused by the collapse of inertial cavitation bubbles, are important for the production of micro- and/or nano-pits. Both the number and diameter of inertial cavitation-induced pits
Y. Tomita et al. / Ultrasonics 55 (2015) 1–5
increased with increasing Tex. The present study suggests that cavitation-induced sonoporation can produce various pore sizes in the membrane, enabling delivery of external molecules of differing sizes into cells or tissues. Acknowledgements The authors would like to express their hearty thanks to Mr. Y. Hamada, SANPICO Ltd., Mr A. Sakamaki, Photron Ltd., and Messrs. S. Tanaka, A. Okada and Y. Takeno for assistance in some of the experiments. Y.T. and T.M. were supported by a JSPS Grant (23656124), and T.K. acknowledges JSPS Grants (26242051) and (24650286). References  S. Bao, B.D. Thrall, D.L. Miller, Transfection of a reporter plasmid into cultured cells by sonoporation in vitro, Ultrasound Med. Biol. 23 (1997) 953–959.  M. Ward, J. Wu, J.-F. Chiu, Ultrasound-induced cell lysis and sonoporation enhanced by contrast agents, J. Acoust. Soc. Am. 105 (1999) 2951–2957.  M.W. Miller, D.L. Miller, A.A. Brayman, A review of in vitro bioeffects of inertial ultrasonic cavitation from a mechanistic perspective, Ultrasound Med. Biol. 22 (1996) 1131–1154.  J. Wu, W.L. Nyborg, Ultrasound, cavitation bubbles and their interaction with cells, Adv. Drug Deliv. Rev. 60 (2008) 1103–1116.  P. Marmottant, S. Hilgenfeldt, Controlled vesicle deformation and lysis by single oscillating bubbles, Nature 423 (2003) 153–156.  P. Prentice, A. Cuschierp, K. Dholakia, M. Prausnitz, P. Campbell, Membrane disruption by optically controlled microbubble cavitation, Nat. Phys. 1 (2005) 107–110.
 M. Postema, A. van Wamel, C.T. Lancee, N. de Jong, Ultrasound-induced encapsulated microbubble phenomena, Ultrasound Med. Biol. 30 (2004) 827–840.  M.M. Forbes, R.L. Steinberg, W.D. O’Brien Jr., Examination of inertial cavitation of Optison™ in producing sonoporation of Chinese hamster ovary cells, Ultrasound Med. Biol. 34 (2008) 2009–2018.  C.F. Caskey, S.M. Stieger, S. Qin, P.A. Dayton, K.W. Ferrara, Direct observations of ultrasound microbubble contrast agent interaction with the microvessel wall, J. Acoust. Soc. Am. 122 (2007) 1191–1200.  H. Chen, A.A. Brayman, W. Kreider, M.R. Bailey, T.J. Matula, Observation of translation and jetting of ultrasound-activated microbubbles in mesenteric microvessels, Ultrasound Med. Biol. 37 (2011) 2139–2148.  J.E. Chomas, P.A. Dayton, M. May, J. Allen, A. Klibanov, K.W. Ferrara, Optical observation of contrast agent destruction, Appl. Phys. Lett. 77 (2000) 1056– 1058.  M. Postema, A. Bouakaz, N. de Jong, Ultrasound-induced gas release from contrast agent microbubbles, EEE Trans Ultrason Ferroelectr. Frequency Control 52 (2005) 1035–1041.  P.C. Sontum, Physicochemical characteristics of Sonazoid™, a new contrast agent for ultrasound imaging, Ultrasound Med. Biol. 34 (2008) 824–833.  Y. Tomita, S. Tanaka, T. Kodama, in: C.D. Ohl, E. Klaseboer, S.W. Ohl, S.W. Gong, B.C. Khoo (Eds.), CAV2012: Proceedings of the 8th International Symposium on Cavitation, Singapore, 14–16 August 2012, CAV2012-Paper No.31.  L. Hoff, Acoustic Characterization of Contrast Agents for Medical Ultrasound Imaging, Kluwer Academic Publishers, Dordrecht, Boston, London, 2001. p. 136.  Y. Tomita, A. Shima, Mechanisms of impulsive pressure generation and damage pit formation by bubble collapse, J. Fluid Mech. 169 (1986) 535–564.  J.R. Blake, G.S. Keen, R.P. Tong, M. Wilson, Acoustic cavitation: the ﬂuid dynamics of non-spherical bubbles, Philos. Trans. R. Soc. Lond. A 357 (1999) 251–267.  S.A. Elder, Cavitation microstreaming, J. Acoust. Soc. Am. 31 (1959) 54–64.  R.T. Knapp, J.W. Daily, F.G. Hammitt, Cavitation, McGraw-Hill, New York, 1970. p. 43.