- Email: [email protected]
Shell thickness determination of polymer-shelled microbubbles using transmission electron microscopy
Accepted Manuscript Title: Shell thickness determination of polymer-shelled microbubbles using transmission electron microscopy Author: Johan H¨armark Hans Hebert Philip J.B. Koeck PII: DOI: Reference:
S0968-4328(16)30044-0 http://dx.doi.org/doi:10.1016/j.micron.2016.03.009 JMIC 2298
To appear in:
Micron
Received date: Revised date: Accepted date:
19-2-2016 31-3-2016 31-3-2016
Please cite this article as: H¨armark, Johan, Hebert, Hans, Koeck, Philip J.B., Shell thickness determination of polymer-shelled microbubbles using transmission electron microscopy.Micron http://dx.doi.org/10.1016/j.micron.2016.03.009 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Shell thickness determination of polymer-shelled microbubbles using transmission electron microscopy
Johan Härmark*, Hans Hebert, Philip J.B. Koeck School of Technology and Health, KTH Royal Institute of Technology and Department of Biosciences and Nutrition, Karolinska Institutet, Stockholm, Sweden
*Corresponding author Johan Härmark School of Technology and Health, KTH Royal Institute of Technology Alfred Nobels Allé 10 141 52 Huddinge, Sweden Telephone number: +46 8 524 810 91 Fax number: + 46 8 21 83 68 [email protected] Email addresses: Philip J.B. Koeck ([email protected]), Hans Hebert ([email protected])
1
Highlights
Introducing a model for shell thickness determination of polymer-shelled MBs The model corrects the measured values in TEM images of cross-sectioned MBs The model was validated using simulated slices of MBs with known dimensions The model can be used for shell thickness determination of micro- and nanoparticles
Abstract Intravenously injected microbubbles (MBs) can be utilized as ultrasound contrast agent (CA) resulting in enhanced image quality. A novel CA, consisting of air filled MBs stabilized with a shell of polyvinyl alcohol (PVA) has been developed. These spherical MBs have been decorated with superparamagnetic iron oxide nanoparticles (SPIONs) in order to serve as both ultrasound and magnetic resonance imaging (MRI) CA. In this study, a mathematical model was introduced that determined the shell thickness of two types of SPIONs decorated MBs (Type A and Type B). The shell thickness of MBs is important to determine, as it affects the acoustical properties. In order to investigate the shell thickness, thin sections of plastic embedded MBs were prepared and imaged using transmission electron microscopy (TEM). However, the sections were cut at random distances from the MB center, which affected the observed shell thickness. Hence, the model determined the average shell thickness of the MBs from corrected mean values of the outer and inner radii observed in the TEM sections. The model was validated using simulated slices of MBs with known shell thickness and radius. The average shell thickness of Type A and Type B MBs were 651 nm and 637 nm, respectively. Keywords: Microbubbles, Shell thickness, Contrast agent, Transmission electron microscopy
2
1. Introduction Intravenously injected solutions containing gas filled bubbles can be utilized as ultrasound contrast agents (CAs) resulting in enhanced image quality (Gramiak and Shah, 1968). The first ultrasound CAs developed for clinical use consisted of gas filled bubbles stabilized by protein molecules (Christiansen et al., 1994; Cohen et al., 1998). The mean diameter of these gas bubbles were typically 2-5 µm (Cohen et al., 1998). Hence, these air-filled CAs are generally referred to as microbubbles (MBs). The first types of MBs were relatively unstable and had short half-life times (Correas et al., 2001). Therefore, other shell materials, such as phospholipids, were used to produce more stable MBs (Morel et al., 2000; Schneider, 1999). However, these MBs have relatively wide size distributions and limited shelf life (Correas et al., 2001; Schneider et al., 1995). In order to produce more rigid CAs, MBs with polymeric shells were developed (Narayan and Wheatley, 1999). One example of this type of CA consists of monodisperse, air filled MBs stabilized with a shell of polyvinyl alcohol (PVA) (Cavalieri et al., 2005). As reported previously, these spherical MBs have been decorated with superparamagnetic iron oxide nanoparticles (SPIONs) in order to serve as both ultrasound and magnetic resonance imaging (MRI) CA (Brismar et al., 2012). Similar strategies have also been investigated using other polymer-shelled MBs (Yang et al., 2009). Additional examples of potential dual modality contrast agents have also been reported recently (Kovalenko et al., 2014; Nguyen et al., 2013; Vlaskou et al., 2010; Zimny et al., 2014). It has been demonstrated that the blood circulation times for polymer-shelled CAs were prolonged compared to phospholipid-shelled CA (du Toit et al., 2011; Härmark et al., 2015). Further modifications of polymer-shelled MBs that combine additional imaging modalities, such as computer tomography and emission imaging, have been investigated (Barrefelt et al., 2013). Finally, polymer-shelled MBs have also been modified to work as a 3
carrier for targeted drug delivery (Gao et al., 2008; Sanna et al., 2011; Villa et al., 2013). In summary, polymer-shelled MBs offer large chemical flexibility and can be used as CA for several imaging modalities, as well as a device for targeted drug delivery. Thus, these polymershelled MBs open up a wide range of possible clinical applications. Further modifications of the polymer-shelled MBs will alter the shell structure. Therefore, the shell thickness can be seen as a compromise between acoustical performance and chemical flexibility. As an example, the MB shell thickness affects the mechanical properties and acoustical behavior, e.g. a thicker and stiffer shell generally reduces the ultrasound signal (Poehlmann et al., 2014). Nevertheless, if the MB shell thickness is known, the ultrasound settings can be selected for optimal signal enhancement. Consequently, the MB shell thickness is an important parameter to determine when introducing additional modifications. The potential of using thin sectioning transmission electron microscopy (TEM) to reveal structural features of gas-filled polymer-shelled particles have been reported elsewhere (Brismar et al., 2012; Harris et al., 1995; Yang et al., 2009). The shell thickness of SPIONs decorated polymer-shelled MBs could be observed in TEM images of plastic embedded, cross-sectioned MBs. However, since the TEM sectioning cuts the MBs at random positions from the MB centers, the average shell thickness observed in the TEM images was overestimated. In this study a mathematical model was introduced that determined the average shell thicknesses of polymer-shelled MBs from corrected values of the outer and inner radius of the MBs.
2. Materials and methods 2.1
Polymer-shelled microbubbles
Two kinds of MBs, previously reported as Type A and Type B (Brismar et al., 2012), were investigated in this study. The shell of these MBs was produced by cross-linking PVA polymers
4
at the air-water interface during high-shear force stirring (Cavalieri et al., 2005). In this process air is encapsulated by a stable polymer shell. Type A was produced by covalent coupling of SPIONs to the PVA-shell structure (Brismar et al., 2012). Type B was produced by adding SPIONs during the PVA cross-linking reaction. Hence, the SPIONs were embedded inside the shell of these MBs (Brismar et al., 2012). The structures of Type A and Type B are illustrated in Figure 1.
2.2
Sample preparation and transmission electron microscopy imaging
In order to determine the shell thickness, thin cross-sections of both Type A and Type B MBs were generated using a previously described protocol (Brismar et al., 2012). In short, MBs were trapped in a gelatin (Merck, Germany) solution and further stabilized by adding paraformaldehyde to the solution. After storing the solution overnight, a pellet was created that was then embedded in epoxy resin, LX 112 (Ladd, USA) following a dehydration and substitution protocol (Hansson et al., 2004). Approximately 50 nm thin sections were cut using a Leica Ultracut UCT (Leica, Austria) and placed on formvar coated, 50 mesh copper grids coated with a thin carbon film. TEM imaging was performed using a Philips/FEI CM120 electron microscope (FEI, The Netherlands) at an acceleration voltage of 120 kV. A total of 100 images of both Type A and Type B MBs, each containing one MB per image, were recorded on a 1k CCD camera at 3,000 x magnification. However, since the MBs were trapped at different levels in the gelatin solution, the apparent shell thickness and radius seen in the images were dependent on the slicing position. Thus, even for MBs with identical radius and shell thickness, sections that were cut at an increasing distance from the MB center will produce larger shell thickness and smaller radius compared to the values at the MB equatorial center, see Figure 2.
5
2.3
Model
In order to improve the estimation of the shell thicknesses of Type A and Type B MBs, a model that corrects for the random slicing effect was introduced. The basis of our model was adapted from models utilized for determining shell thicknesses of microcapsules and yeast cells (Mercadé-Prieto et al., 2011; Smith et al., 2000). In those models, the observed outer radius, shell thickness and slicing angle, are integrated within a range of angles where the apparent inner radius is different from 0, i.e. excluding images at very high slicing angles that only cuts through the shell. Hence, a limitation with those models is that they only allow images where both the outer and inner radius can be observed. To use these models, the limits for the slicing angle first need to be determined numerically in order to calculate the shell thickness. Instead of using this approach, we introduce a model for determining the MB shell thickness directly from the TEM images by independently calculating the mean outer and inner radius and then taking the difference of these as the average MB shell thickness. Thus, using our model, as described in the following section, the limits of the slicing angles are known. Following the definitions in Figure 3, the projection z along the y-axis at slicing angle φ can be expressed as 𝑧 = 𝑅 sin 𝜑
→
𝑧
𝜑 = sin−1 𝑅
(1)
The projection z is limited to slicing at heights ranging from –R to +R, while cuts at a position within this range along z should be of equal probability. Therefore, the probability function for cuts along the projection z can be written as 1
𝑓(𝑧) = 2𝑅 , for |𝑧| ≤ 𝑅, else 0
→
+𝑅
1
∫−𝑅 𝑓(𝑧)𝑑𝑧 = 2𝑅 ∙ [𝑧]+𝑅 −𝑅 = 1
6
(2)
The observed outer radius (Ro) is given by geometry 𝑅𝑜 (𝑧) = 𝑅 cos(𝜑(𝑧))
(3)
The mean value of the observed outer radius (𝑅̅o) can be expressed by the following expectation value, where f(R) is the probability distribution of the outer radius ∞ ∞ 𝑅̅𝑜 = ∫0 𝑓(𝑅) ∫−∞ 𝑅 cos 𝜑 𝑓 (𝑧) 𝑑𝑧 𝑑𝑅
(4)
In this case the projection z varies between –R and +R. Thus, equation (4) can be written as ∞ 𝑅 𝑅̅𝑜 = ∫0 𝑓(𝑅) ∫−𝑅 𝑅 cos 𝜑
1 2𝑅
𝑑𝑧 𝑑𝑅
(5)
Using equation (1), dz can be expressed 𝑑𝑧 = 𝑅 cos 𝜑 𝑑𝜑,
with
𝜋
−2 ≤𝜑 ≤
𝜋 2
(6)
Replacing parameters for z in equation (5) using equation (6) this expression becomes ∞ 𝜋/2 1 𝑅̅𝑜 = ∫0 𝑓(𝑅) ∫–𝜋/2 𝑅 cos 𝜑 2𝑅 𝑅 cos 𝜑 𝑑𝜑 𝑑𝑅
(7)
Finally, the integral for R can be replaced by its expectation value 𝑅̅, giving this equation 1 𝜋/2 𝜋 𝑅̅𝑜 = 𝑅̅ 2 ∫–𝜋/2 cos 2 𝜑 𝑑𝜑 = 4 ∙ 𝑅̅
(8)
̅ ) can be calculated from the mean observed outer As seen in equation (8) the mean outer radius (R ̅ o ). radius (R The mean inner radius (r̅ ) can be calculated from the mean observed inner radius (r̅ o ) using the same methodology as for the outer radius, see Figure 3
7
𝑧 = 𝑟 sin 𝜃
→
𝑧
𝜃 = sin−1 𝑟
(9)
The projection z now ranging from –r to +r and within this range, cuts along z should be of equal probability 1
𝑓 (𝑧) = 2𝑟 , for |𝑧| ≤ 𝑟, else 0
→
+𝑟
1
∫−𝑟 𝑓(𝑧)𝑑𝑧 = 2𝑟 ∙ [𝑧]+𝑟 −𝑟 = 1
(10)
The observed inner radius (ro) is given by geometry 𝑟𝑜 (𝑧) = 𝑟 cos(𝜃 (𝑧))
(11)
In analogy to the outer radius, the mean observed inner radius (ro) can be determined by solving these integrals, where f(r) is the probability distribution of the inner radius ∞
∞
𝜋
𝑟̅𝑜 = ∫0 𝑓(𝑟) ∫−∞ 𝑟 cos 𝜃 𝑓 (𝑧) 𝑑𝑧 𝑑𝑟 = 4 ∙ 𝑟̅
(12)
̅ can then be calculated from the experimental data using equation (8) The mean shell thickness T and (12) 4 𝑇̅ = 𝑅̅ − 𝑟̅ = 𝜋 (𝑅̅𝑜 − 𝑟̅𝑜 )
(13)
Since the inner radius does not exist for slices that only cut through the MB shell, e.g. for projections (z), r < z ≤ R, the data sets for calculating the mean inner and outer radius will have different size.
2.4
Simulation
In order to validate this model, simulated slices of MBs with known shell thickness and radius were generated using Microsoft Excel (2007). For each selected MB radius (R) (1500, 2000 and 2500 nm) the shell thickness (T) was set to 200, 400, 600, 800 and 1000 nm. Thus, altogether 15 8
MBs with varying radius and shell thickness were simulated. For each MB dimension, a total of 1000 sections were generated computationally by slicing the MB at random projections z ranging from –R to R. For each slice, the observed outer radius (Ro) and inner radius (ro) were derived by basic geometry, see Figure 3. By using equation (13) the mean shell thickness of the simulated MBs were calculated. Finally, the simulation was iterated 10 times for each selected MB dimension.
3. Results and discussion 3.1
Model for determining the microbubble shell thickness
The mean shell thicknesses of two polymer-based MBs were calculated from TEM images; containing cross-sections of the MBs, using the mathematical model described herein, see Table ̅) of MBs, with SPIONs covalently attached to the PVA-shell 1. The average shell thickness (T (Type A), was 651 nm. For Type B MBs, with SPIONs embedded inside the shell, the average ̅) was 637 nm. In addition, the average radius of Type A MBs (R ̅ = 1882 nm) shell thickness (T ̅ = 1809 nm). was larger than Type B MBs (R
3.2
Simulation of microbubbles with known dimensions
Previous studies, using other techniques and models for measuring the average shell thickness of polymer-shelled MBs, have reported shell thickness values ranging from 200-900 nm (Cavalieri et al., 2005; Poehlmann et al., 2014). Therefore, slices of MBs with known shell thickness and radius were generated computationally to validate the model. The mean shell thickness and standard deviation from 10 repeated MB simulations, where each simulation included 1000 sections, are listed in Table 2.
9
3.3
Impact of microbubble shell thickness
As seen in Table 2, the mean shell thicknesses calculated from the model correlates with the known values of the MBs. Thus, these simulations show that the model presented in this paper can be used for accurately determining the average shell thickness of MBs. The shell thickness is an important parameter as it affects the acoustical properties of the MBs. Knowing the average shell thickness of the MBs, the settings for optimizing the signal-to-noise ratio during ultrasound imaging can be determined. As reported in this study, the shell of Type A MBs was on average 14 nm thicker than the shell of Type B MBs, see Table 1. These results were expected since the SPIONs (diameter 8-12 nm (Mikhaylova et al., 2004)) were defined as part of the MB shell. In addition, Type A MBs are softer compared to Type B MBs, due to the fact that the SPIONs embedded inside the shell are stabilizing the PVA network (Poehlmann et al., 2014). Hence, even though the Type A MBs are loaded with higher amount of SPIONs (Brismar et al., 2012), comparable shell thicknesses could be explained by the fact that the network of PVA was slightly compressed in Type A, when the SPIONs were loaded on the surface. Although the ultrasound response is decreased for thick and rigid shelled MBs, previous studies have shown that stable polymer-shelled MBs can be used as a multimodal CA and as a carrier for targeted drug delivery (Barrefelt et al., 2013; Brismar et al., 2012; Villa et al., 2013). To further investigate the shell wall structure of Type B MBs, an equation that estimates the radial distribution of SPIONs from TEM images could potentially be employed (Mai and Eisenberg, 2010). This information could be useful for optimizing the MRI contrast enhancement capability of these MBs.
3.4
Limitations of the model
The model described herein is only valid for spherical objects. In this study the analyzed samples were rigid objects, observed as spherical structures in the TEM sections. However, a few 10
collapsed MBs, mainly Type A, were observed in the TEM sections. Although none of these MBs were included in the study. During TEM imaging, it is possible that cuts at projections z close to -R or +R, were missed due to their small size and therefore not included in the analysis. ̅ and thus also the average shell thickness could be overestimated. Another As a result, R limitation of the model was that no information on the variation of the MB shell thickness was obtained. Still, determining the average MB shell thickness is very useful for selecting the optimal imaging conditions. As shown in simulations using MBs with known shell thickness and radius, the difference between the true and calculated shell thickness was within the range of 1030 nm.
3.5
Other applications for the model
The model described in this study has been used for determining the shell thickness of two types of MBs. Theoretically, this model could also be applied for determining the shell thickness of other types of MBs. Other applications for using this model could be for estimating the cell-wall thicknesses of biological samples or shell thicknesses of microcapsules and other spherical structures. In the emerging field of nanotechnology, parameters of pharmaceutical substances, such as the shell thickness, need to be evaluated before regulatory approval. Since images can be recorded at high magnification using TEM, this model can in principle be used for determining shell thicknesses of such nanometer-sized substances.
11
4. Conclusions This study shows that the average shell thickness of polymer-shelled MBs could be determined from TEM images of cross-sectioned MBs using the mathematical model described herein. The model was validated using simulated slices of MBs with known shell thickness and radius and the calculated average shell thickness correlated with the known values. The average shell thickness of Type A MBs, with SPIONs covalently attached to the MB shell surface, was 651 nm. Whereas the average shell thickness of Type B MBs, with SPIONs embedded inside the shell, was 637 nm. This model could potentially be used for determining shell thicknesses of microcapsules and nanometer-sized pharmaceutical substances.
Acknowledgements This study was a part of the 3MiCRON (245572) project, which was founded by the European Commission within the Seventh Framework Program. We thank Dr Kjell Hultenby and his staff at the Electron Microscopy Unit (EMil) at the Karolinska University Hospital (Huddinge, Sweden) for preparing the MB thin sections. We also thank Prof. Gaio Paradossi and his colleagues at the Department of Chemical Sciences and Technologies at the Università di Roma Tor Vergata (Rome, Italy) for preparing Type A and Type B MBs. Support from the Center for Innovative Medicine (CIMED), Karolinska Institutet (Huddinge, Sweden) is also acknowledged.
12
References Barrefelt, Å.A., Brismar, T.B., Egri, G., Aspelin, P., Olsson, A., Oddo, L., Margheritelli, S., Caidahl, K., Paradossi, G., Dähne, L., Barrefelt, A., Brismar, T.B., Egri, G., Aspelin, P., Olsson, A., Oddo, L., Margheritelli, S., Caidahl, K., Paradossi, G., Dahne, L., Axelsson, R., Hassan, M., 2013. Multimodality imaging using SPECT/CT and MRI and ligand functionalized 99mTc-labeled magnetic microbubbles. EJNMMI Res. 3, 12. doi:10.1186/2191-219X-3-12 Brismar, T.B., Grishenkov, D., Gustafsson, B., Härmark, J., Barrefelt, A., Kothapalli, S.V.V.N., Margheritelli, S., Oddo, L., Caidahl, K., Hebert, H., Paradossi, G., Harmark, J., Barrefelt, A., Kothapalli, S.V.V.N., Margheritelli, S., Oddo, L., Caidahl, K., Hebert, H., Paradossi, G., 2012. Magnetite nanoparticles can be coupled to microbubbles to support multimodal imaging. Biomacromolecules 13, 1390–1399. doi:10.1021/bm300099f Cavalieri, F., El Hamassi, A., Chiessi, E., Paradossi, G., 2005. Stable polymeric microballoons as multifunctional device for biomedical uses: Synthesis and characterization. Langmuir 21, 8758–8764. doi:10.1021/la050287j Christiansen, C., Kryvi, H., Sontum, P., Skotland, T., 1994. Physical and biochemical characterization of Albunex, a new ultrasound contrast agent consisting of air-filled albumin microspheres suspended in a solution of human albumin. Biotechnol. Appl. Biochem. 19, 307–320. doi:10.1111/j.1470-8744.1994.tb00300.x Cohen, J.L., Cheirif, J., Segar, D.S., Gillam, L.D., Gottdiener, J.S., Hausnerova, E., Bruns, D.E., 1998. Improved left ventricular endocardial border delineation and opacification with Optison (FS069), a new echocardiographic contrast agent: results of a phase III multicenter trial. J. Am. Coll. Cardiol. 32, 746–752. doi:10.1016/S0735-1097(98)00311-8 Correas, J.-M., Bridal, L., Lesavre, A., Méjean, A., Claudon, M., Hélénon, O., 2001. Ultrasound contrast agents: properties, principles of action, tolerance, and artifacts. Eur. Radiol. 11, 1316–1328. doi:10.1007/s003300100940 du Toit, L.C., Govender, T., Pillay, V., Choonara, Y.E., Kodama, T., 2011. Investigating the Effect of Polymeric Approaches on Circulation Time and Physical Properties of Nanobubbles. Pharm. Res. 28, 494–504. doi:10.1007/s11095-010-0247-y Gao, Z., Kennedy, A.M., Christensen, D.A., Rapoport, N.Y., 2008. Drug-loaded nano/microbubbles for combining ultrasonography and targeted chemotherapy. Ultrasonics 48, 260–270. doi:10.1016/j.ultras.2007.11.002 Gramiak, R., Shah, P.M., 1968. Echocardiography of the aortic root. Invest. Radiol. 3, 356–366. Hansson, A., Hance, N., Dufour, E., Rantanen, A., Hultenby, K., Clayton, D.A., Wibom, R., Larsson, N.-G.G., 2004. A switch in metabolism precedes increased mitochondrial biogenesis in respiratory chain-deficient mouse hearts. Proc. Natl. Acad. Sci. U. S. A. 101, 3136–3141. doi:10.1073/pnas.0308710100 Harris, J.R., Depoix, F., Urich, K., 1995. The structure of gas-filled n-butyl-2-cyanoacrylate 13
(BCA) polymer particles. Micron 26, 103–111. doi:10.1016/0968-4328(95)93275-H Härmark, J., Larsson, M.K., Razuvajev, A., Koeck, P.J.B., Paradossi, G., Brodin, L.-Å., Caidahl, K., Hebert, H., Bjällmark, A., 2015. Investigation of the elimination process of a multimodal polymer-shelled contrast agent in rats using ultrasound and transmission electron microscopy. Biomed. Spectrosc. Imaging 4, 81–93. doi:10.3233/BSI-140099 Kovalenko, A., Jouhannaud, J., Polavarapu, P., Krafft, M.P., Waton, G., Pourroy, G., 2014. Hollow magnetic microspheres obtained by nanoparticle adsorption on surfactant stabilized microbubbles. Soft Matter 10, 5147–5156. doi:10.1039/c4sm00318g Mai, Y., Eisenberg, A., 2010. Controlled Incorporation of Particles into the Central Portion of Vesicle Walls. J. Am. Chem. Soc. 132, 10078–10084. doi:10.1021/ja1024063 Mercadé-Prieto, R., Nguyen, B., Allen, R., York, D., Preece, J.A., Goodwin, T.E., Zhang, Z., 2011. Determination of the elastic properties of single microcapsules using micromanipulation and finite element modeling. Chem. Eng. Sci. 66, 2042–2049. doi:10.1016/j.ces.2011.01.015 Mikhaylova, M., Kim, D.K., Berry, C.C., Zagorodni, A., Toprak, M., Curtis, A.S.G., Muhammed, M., 2004. BSA immobilization on amine-functionalized superparamagnetic iron oxide nanoparticles. Chem. Mater. 16, 2344–2354. doi:10.1021/cm0348904 Morel, D.R., Schwieger, I., Hohn, L., Terrettaz, J., Llull, J.B., Cornioley, Y.A., Schneider, M., 2000. Human Pharmacokinetics and Safety Evaluation of SonoVueTM, a New Contrast Agent for Ultrasound Imaging. Invest. Radiol. 35, 80–5. doi:10.1097/00004424-20000100000009 Narayan, P., Wheatley, M.A., 1999. Preparation and characterization of hollow microcapsules for use as ultrasound contrast agents. Polym. Eng. Sci. 39, 2242–2255. doi:10.1002/pen.11612 Nguyen, P.N., Nikolova, G., Polavarapu, P., Waton, G., Phuoc, L.T., Pourroy, G., Krafft, M.P., 2013. Compressible multi-scale magnetic constructs: decorating the outer surface of selfassembled microbubbles with iron oxide nanoparticles. RSC Adv. 3, 7743–7746. doi:10.1039/c3ra40908b Poehlmann, M., Grishenkov, D., Kothapalli, S.V.V.N., Härmark, J., Hebert, H., Philipp, A., Hoeller, R., Seuss, M., Kuttner, C., Margheritelli, S., Paradossi, G., Fery, A., 2014. On the interplay of shell structure with low- and high-frequency mechanics of multifunctional magnetic microbubbles. Soft Matter 10, 214–226. doi:10.1039/C3SM51560E Sanna, V., Pintus, G., Bandiera, P., Anedda, R., Punzoni, S., Sanna, B., Migaleddu, V., Uzzau, S., Sechi, M., 2011. Development of Polymeric Microbubbles Targeted to Prostate-Specific Membrane Antigen as Prototype of Novel Ultrasound Contrast Agents. Mol. Pharm. 8, 748– 757. doi:10.1021/mp100360g Schneider, M., 1999. SonoVue, a new ultrasound contrast agent. Eur. Radiol. 9, S347–S348. doi:10.1007/pl00014071 Schneider, M., Arditi, M., Barrau, M.-B., Brochot, J., Broillet, A., Ventrone, R., Yan, F., 1995. 14
BR1: a new ultrasonographic contrast agent based on sulfur hexafluoride-filled microbubbles. Invest. Radiol. 30, 451–457. Smith, A.E., Zhang, Z., Thomas, C.R., 2000. Wall material properties of yeast cells: Part 1. Cell measurements and compression experiments. Chem. Eng. Sci. 55, 2031–2041. doi:10.1016/S0009-2509(99)00500-X Villa, R., Cerroni, B., Viganò, L., Margheritelli, S., Abolafio, G., Oddo, L., Paradossi, G., Zaffaroni, N., 2013. Targeted doxorubicin delivery by chitosan-galactosylated modified polymer microbubbles to hepatocarcinoma cells. Colloids Surfaces B Biointerfaces 110, 434–442. doi:10.1016/j.colsurfb.2013.04.022 Vlaskou, D., Mykhaylyk, O., Krötz, F., Hellwig, N., Renner, R., Schillinger, U., Gleich, B., Heidsieck, A., Schmitz, G., Hensel, K., Plank, C., 2010. Magnetic and Acoustically Active Lipospheres for Magnetically Targeted Nucleic Acid Delivery. Adv. Funct. Mater. 20, 3881–3894. doi:10.1002/adfm.200902388 Yang, F., Li, Y.X., Chen, Z.P., Zhang, Y., Wu, J.R., Gu, N., 2009. Superparamagnetic iron oxide nanoparticle-embedded encapsulated microbubbles as dual contrast agents of magnetic resonance and ultrasound imaging. Biomaterials 30, 3882–3890. doi:10.1016/j.biomaterials.2009.03.051 Zimny, K., Mascaro, B., Brunet, T., Poncelet, O., Aristégui, C., Leng, J., Sandre, O., MondainMonval, O., 2014. Design of a fluorinated magneto-responsive material with tuneable ultrasound scattering properties. J. Mater. Chem. B 2, 1285–1297. doi:10.1039/c3tb21585g
15
Figure 1 Transmission electron microscopy images of a Type A (A) and Type B (B) microbubbles. The dark particles (arrows) are superparamagnetic iron oxide nanoparticles attached to (A), or embedded in (B), the polymer shell. Scale bars represent 500 nm.
Figure 2 A theoretical section through the epoxy resin (grey). As illustrated, slicing at varying distances from the MB center will affect the observed shell thickness (T1 > T2).
Figure 3 Illustrating the parameters used in our model for determining the shell thickness of spherical objects. T is defined as the shell thickness, R as the outer radius, r as the inner radius, Ro as the observed outer radius, ro as the observed inner radius, z as the projected y-axis height at slicing angles φ and θ and, finally To as the observed shell thickness.
16
Fig.1
17
Fig.2
18
Fig.3
19
̅ of two types of microbubbles determined by the model Table 1 Average shell thickness 𝐓
Microbubbles
̅ 𝐨 [nm] 𝐑
𝐫̅𝐨 [nm]
̅ [nm] 𝐑
𝐫̅ [nm]
̅ [nm] 𝐓
Type A
1478
967
1882
1231
651
Type B
1421
920
1809
1171
637
̅) and inner (r̅) radius were calculated from the mean observed outer (R ̅ o ) (n=100 for Type A and Type B) Notes: the mean outer (R and inner (r̅o) (n=80 for Type A and n=82 for Type B) radius of the microbubbles in the transmission electron microscopy sections.
20
Table 2 The average shell thickness and standard deviation calculated from the model by iteratively (n = 10) simulate cutting random slices (n = 1000) of microbubbles with known shell thickness and radius.
Shell thickness [nm] Radius [nm]
200
400
600
800
1000
1500
204±10
394±11
601±16
800±14
1000±15
2000
204±10
404±13
602±15
795±15
996±17
2500
204±14
405±17
607±27
800±24
993±34
21
S0968-4328(16)30044-0 http://dx.doi.org/doi:10.1016/j.micron.2016.03.009 JMIC 2298
To appear in:
Micron
Received date: Revised date: Accepted date:
19-2-2016 31-3-2016 31-3-2016
Please cite this article as: H¨armark, Johan, Hebert, Hans, Koeck, Philip J.B., Shell thickness determination of polymer-shelled microbubbles using transmission electron microscopy.Micron http://dx.doi.org/10.1016/j.micron.2016.03.009 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Shell thickness determination of polymer-shelled microbubbles using transmission electron microscopy
Johan Härmark*, Hans Hebert, Philip J.B. Koeck School of Technology and Health, KTH Royal Institute of Technology and Department of Biosciences and Nutrition, Karolinska Institutet, Stockholm, Sweden
*Corresponding author Johan Härmark School of Technology and Health, KTH Royal Institute of Technology Alfred Nobels Allé 10 141 52 Huddinge, Sweden Telephone number: +46 8 524 810 91 Fax number: + 46 8 21 83 68 [email protected] Email addresses: Philip J.B. Koeck ([email protected]), Hans Hebert ([email protected])
1
Highlights
Introducing a model for shell thickness determination of polymer-shelled MBs The model corrects the measured values in TEM images of cross-sectioned MBs The model was validated using simulated slices of MBs with known dimensions The model can be used for shell thickness determination of micro- and nanoparticles
Abstract Intravenously injected microbubbles (MBs) can be utilized as ultrasound contrast agent (CA) resulting in enhanced image quality. A novel CA, consisting of air filled MBs stabilized with a shell of polyvinyl alcohol (PVA) has been developed. These spherical MBs have been decorated with superparamagnetic iron oxide nanoparticles (SPIONs) in order to serve as both ultrasound and magnetic resonance imaging (MRI) CA. In this study, a mathematical model was introduced that determined the shell thickness of two types of SPIONs decorated MBs (Type A and Type B). The shell thickness of MBs is important to determine, as it affects the acoustical properties. In order to investigate the shell thickness, thin sections of plastic embedded MBs were prepared and imaged using transmission electron microscopy (TEM). However, the sections were cut at random distances from the MB center, which affected the observed shell thickness. Hence, the model determined the average shell thickness of the MBs from corrected mean values of the outer and inner radii observed in the TEM sections. The model was validated using simulated slices of MBs with known shell thickness and radius. The average shell thickness of Type A and Type B MBs were 651 nm and 637 nm, respectively. Keywords: Microbubbles, Shell thickness, Contrast agent, Transmission electron microscopy
2
1. Introduction Intravenously injected solutions containing gas filled bubbles can be utilized as ultrasound contrast agents (CAs) resulting in enhanced image quality (Gramiak and Shah, 1968). The first ultrasound CAs developed for clinical use consisted of gas filled bubbles stabilized by protein molecules (Christiansen et al., 1994; Cohen et al., 1998). The mean diameter of these gas bubbles were typically 2-5 µm (Cohen et al., 1998). Hence, these air-filled CAs are generally referred to as microbubbles (MBs). The first types of MBs were relatively unstable and had short half-life times (Correas et al., 2001). Therefore, other shell materials, such as phospholipids, were used to produce more stable MBs (Morel et al., 2000; Schneider, 1999). However, these MBs have relatively wide size distributions and limited shelf life (Correas et al., 2001; Schneider et al., 1995). In order to produce more rigid CAs, MBs with polymeric shells were developed (Narayan and Wheatley, 1999). One example of this type of CA consists of monodisperse, air filled MBs stabilized with a shell of polyvinyl alcohol (PVA) (Cavalieri et al., 2005). As reported previously, these spherical MBs have been decorated with superparamagnetic iron oxide nanoparticles (SPIONs) in order to serve as both ultrasound and magnetic resonance imaging (MRI) CA (Brismar et al., 2012). Similar strategies have also been investigated using other polymer-shelled MBs (Yang et al., 2009). Additional examples of potential dual modality contrast agents have also been reported recently (Kovalenko et al., 2014; Nguyen et al., 2013; Vlaskou et al., 2010; Zimny et al., 2014). It has been demonstrated that the blood circulation times for polymer-shelled CAs were prolonged compared to phospholipid-shelled CA (du Toit et al., 2011; Härmark et al., 2015). Further modifications of polymer-shelled MBs that combine additional imaging modalities, such as computer tomography and emission imaging, have been investigated (Barrefelt et al., 2013). Finally, polymer-shelled MBs have also been modified to work as a 3
carrier for targeted drug delivery (Gao et al., 2008; Sanna et al., 2011; Villa et al., 2013). In summary, polymer-shelled MBs offer large chemical flexibility and can be used as CA for several imaging modalities, as well as a device for targeted drug delivery. Thus, these polymershelled MBs open up a wide range of possible clinical applications. Further modifications of the polymer-shelled MBs will alter the shell structure. Therefore, the shell thickness can be seen as a compromise between acoustical performance and chemical flexibility. As an example, the MB shell thickness affects the mechanical properties and acoustical behavior, e.g. a thicker and stiffer shell generally reduces the ultrasound signal (Poehlmann et al., 2014). Nevertheless, if the MB shell thickness is known, the ultrasound settings can be selected for optimal signal enhancement. Consequently, the MB shell thickness is an important parameter to determine when introducing additional modifications. The potential of using thin sectioning transmission electron microscopy (TEM) to reveal structural features of gas-filled polymer-shelled particles have been reported elsewhere (Brismar et al., 2012; Harris et al., 1995; Yang et al., 2009). The shell thickness of SPIONs decorated polymer-shelled MBs could be observed in TEM images of plastic embedded, cross-sectioned MBs. However, since the TEM sectioning cuts the MBs at random positions from the MB centers, the average shell thickness observed in the TEM images was overestimated. In this study a mathematical model was introduced that determined the average shell thicknesses of polymer-shelled MBs from corrected values of the outer and inner radius of the MBs.
2. Materials and methods 2.1
Polymer-shelled microbubbles
Two kinds of MBs, previously reported as Type A and Type B (Brismar et al., 2012), were investigated in this study. The shell of these MBs was produced by cross-linking PVA polymers
4
at the air-water interface during high-shear force stirring (Cavalieri et al., 2005). In this process air is encapsulated by a stable polymer shell. Type A was produced by covalent coupling of SPIONs to the PVA-shell structure (Brismar et al., 2012). Type B was produced by adding SPIONs during the PVA cross-linking reaction. Hence, the SPIONs were embedded inside the shell of these MBs (Brismar et al., 2012). The structures of Type A and Type B are illustrated in Figure 1.
2.2
Sample preparation and transmission electron microscopy imaging
In order to determine the shell thickness, thin cross-sections of both Type A and Type B MBs were generated using a previously described protocol (Brismar et al., 2012). In short, MBs were trapped in a gelatin (Merck, Germany) solution and further stabilized by adding paraformaldehyde to the solution. After storing the solution overnight, a pellet was created that was then embedded in epoxy resin, LX 112 (Ladd, USA) following a dehydration and substitution protocol (Hansson et al., 2004). Approximately 50 nm thin sections were cut using a Leica Ultracut UCT (Leica, Austria) and placed on formvar coated, 50 mesh copper grids coated with a thin carbon film. TEM imaging was performed using a Philips/FEI CM120 electron microscope (FEI, The Netherlands) at an acceleration voltage of 120 kV. A total of 100 images of both Type A and Type B MBs, each containing one MB per image, were recorded on a 1k CCD camera at 3,000 x magnification. However, since the MBs were trapped at different levels in the gelatin solution, the apparent shell thickness and radius seen in the images were dependent on the slicing position. Thus, even for MBs with identical radius and shell thickness, sections that were cut at an increasing distance from the MB center will produce larger shell thickness and smaller radius compared to the values at the MB equatorial center, see Figure 2.
5
2.3
Model
In order to improve the estimation of the shell thicknesses of Type A and Type B MBs, a model that corrects for the random slicing effect was introduced. The basis of our model was adapted from models utilized for determining shell thicknesses of microcapsules and yeast cells (Mercadé-Prieto et al., 2011; Smith et al., 2000). In those models, the observed outer radius, shell thickness and slicing angle, are integrated within a range of angles where the apparent inner radius is different from 0, i.e. excluding images at very high slicing angles that only cuts through the shell. Hence, a limitation with those models is that they only allow images where both the outer and inner radius can be observed. To use these models, the limits for the slicing angle first need to be determined numerically in order to calculate the shell thickness. Instead of using this approach, we introduce a model for determining the MB shell thickness directly from the TEM images by independently calculating the mean outer and inner radius and then taking the difference of these as the average MB shell thickness. Thus, using our model, as described in the following section, the limits of the slicing angles are known. Following the definitions in Figure 3, the projection z along the y-axis at slicing angle φ can be expressed as 𝑧 = 𝑅 sin 𝜑
→
𝑧
𝜑 = sin−1 𝑅
(1)
The projection z is limited to slicing at heights ranging from –R to +R, while cuts at a position within this range along z should be of equal probability. Therefore, the probability function for cuts along the projection z can be written as 1
𝑓(𝑧) = 2𝑅 , for |𝑧| ≤ 𝑅, else 0
→
+𝑅
1
∫−𝑅 𝑓(𝑧)𝑑𝑧 = 2𝑅 ∙ [𝑧]+𝑅 −𝑅 = 1
6
(2)
The observed outer radius (Ro) is given by geometry 𝑅𝑜 (𝑧) = 𝑅 cos(𝜑(𝑧))
(3)
The mean value of the observed outer radius (𝑅̅o) can be expressed by the following expectation value, where f(R) is the probability distribution of the outer radius ∞ ∞ 𝑅̅𝑜 = ∫0 𝑓(𝑅) ∫−∞ 𝑅 cos 𝜑 𝑓 (𝑧) 𝑑𝑧 𝑑𝑅
(4)
In this case the projection z varies between –R and +R. Thus, equation (4) can be written as ∞ 𝑅 𝑅̅𝑜 = ∫0 𝑓(𝑅) ∫−𝑅 𝑅 cos 𝜑
1 2𝑅
𝑑𝑧 𝑑𝑅
(5)
Using equation (1), dz can be expressed 𝑑𝑧 = 𝑅 cos 𝜑 𝑑𝜑,
with
𝜋
−2 ≤𝜑 ≤
𝜋 2
(6)
Replacing parameters for z in equation (5) using equation (6) this expression becomes ∞ 𝜋/2 1 𝑅̅𝑜 = ∫0 𝑓(𝑅) ∫–𝜋/2 𝑅 cos 𝜑 2𝑅 𝑅 cos 𝜑 𝑑𝜑 𝑑𝑅
(7)
Finally, the integral for R can be replaced by its expectation value 𝑅̅, giving this equation 1 𝜋/2 𝜋 𝑅̅𝑜 = 𝑅̅ 2 ∫–𝜋/2 cos 2 𝜑 𝑑𝜑 = 4 ∙ 𝑅̅
(8)
̅ ) can be calculated from the mean observed outer As seen in equation (8) the mean outer radius (R ̅ o ). radius (R The mean inner radius (r̅ ) can be calculated from the mean observed inner radius (r̅ o ) using the same methodology as for the outer radius, see Figure 3
7
𝑧 = 𝑟 sin 𝜃
→
𝑧
𝜃 = sin−1 𝑟
(9)
The projection z now ranging from –r to +r and within this range, cuts along z should be of equal probability 1
𝑓 (𝑧) = 2𝑟 , for |𝑧| ≤ 𝑟, else 0
→
+𝑟
1
∫−𝑟 𝑓(𝑧)𝑑𝑧 = 2𝑟 ∙ [𝑧]+𝑟 −𝑟 = 1
(10)
The observed inner radius (ro) is given by geometry 𝑟𝑜 (𝑧) = 𝑟 cos(𝜃 (𝑧))
(11)
In analogy to the outer radius, the mean observed inner radius (ro) can be determined by solving these integrals, where f(r) is the probability distribution of the inner radius ∞
∞
𝜋
𝑟̅𝑜 = ∫0 𝑓(𝑟) ∫−∞ 𝑟 cos 𝜃 𝑓 (𝑧) 𝑑𝑧 𝑑𝑟 = 4 ∙ 𝑟̅
(12)
̅ can then be calculated from the experimental data using equation (8) The mean shell thickness T and (12) 4 𝑇̅ = 𝑅̅ − 𝑟̅ = 𝜋 (𝑅̅𝑜 − 𝑟̅𝑜 )
(13)
Since the inner radius does not exist for slices that only cut through the MB shell, e.g. for projections (z), r < z ≤ R, the data sets for calculating the mean inner and outer radius will have different size.
2.4
Simulation
In order to validate this model, simulated slices of MBs with known shell thickness and radius were generated using Microsoft Excel (2007). For each selected MB radius (R) (1500, 2000 and 2500 nm) the shell thickness (T) was set to 200, 400, 600, 800 and 1000 nm. Thus, altogether 15 8
MBs with varying radius and shell thickness were simulated. For each MB dimension, a total of 1000 sections were generated computationally by slicing the MB at random projections z ranging from –R to R. For each slice, the observed outer radius (Ro) and inner radius (ro) were derived by basic geometry, see Figure 3. By using equation (13) the mean shell thickness of the simulated MBs were calculated. Finally, the simulation was iterated 10 times for each selected MB dimension.
3. Results and discussion 3.1
Model for determining the microbubble shell thickness
The mean shell thicknesses of two polymer-based MBs were calculated from TEM images; containing cross-sections of the MBs, using the mathematical model described herein, see Table ̅) of MBs, with SPIONs covalently attached to the PVA-shell 1. The average shell thickness (T (Type A), was 651 nm. For Type B MBs, with SPIONs embedded inside the shell, the average ̅) was 637 nm. In addition, the average radius of Type A MBs (R ̅ = 1882 nm) shell thickness (T ̅ = 1809 nm). was larger than Type B MBs (R
3.2
Simulation of microbubbles with known dimensions
Previous studies, using other techniques and models for measuring the average shell thickness of polymer-shelled MBs, have reported shell thickness values ranging from 200-900 nm (Cavalieri et al., 2005; Poehlmann et al., 2014). Therefore, slices of MBs with known shell thickness and radius were generated computationally to validate the model. The mean shell thickness and standard deviation from 10 repeated MB simulations, where each simulation included 1000 sections, are listed in Table 2.
9
3.3
Impact of microbubble shell thickness
As seen in Table 2, the mean shell thicknesses calculated from the model correlates with the known values of the MBs. Thus, these simulations show that the model presented in this paper can be used for accurately determining the average shell thickness of MBs. The shell thickness is an important parameter as it affects the acoustical properties of the MBs. Knowing the average shell thickness of the MBs, the settings for optimizing the signal-to-noise ratio during ultrasound imaging can be determined. As reported in this study, the shell of Type A MBs was on average 14 nm thicker than the shell of Type B MBs, see Table 1. These results were expected since the SPIONs (diameter 8-12 nm (Mikhaylova et al., 2004)) were defined as part of the MB shell. In addition, Type A MBs are softer compared to Type B MBs, due to the fact that the SPIONs embedded inside the shell are stabilizing the PVA network (Poehlmann et al., 2014). Hence, even though the Type A MBs are loaded with higher amount of SPIONs (Brismar et al., 2012), comparable shell thicknesses could be explained by the fact that the network of PVA was slightly compressed in Type A, when the SPIONs were loaded on the surface. Although the ultrasound response is decreased for thick and rigid shelled MBs, previous studies have shown that stable polymer-shelled MBs can be used as a multimodal CA and as a carrier for targeted drug delivery (Barrefelt et al., 2013; Brismar et al., 2012; Villa et al., 2013). To further investigate the shell wall structure of Type B MBs, an equation that estimates the radial distribution of SPIONs from TEM images could potentially be employed (Mai and Eisenberg, 2010). This information could be useful for optimizing the MRI contrast enhancement capability of these MBs.
3.4
Limitations of the model
The model described herein is only valid for spherical objects. In this study the analyzed samples were rigid objects, observed as spherical structures in the TEM sections. However, a few 10
collapsed MBs, mainly Type A, were observed in the TEM sections. Although none of these MBs were included in the study. During TEM imaging, it is possible that cuts at projections z close to -R or +R, were missed due to their small size and therefore not included in the analysis. ̅ and thus also the average shell thickness could be overestimated. Another As a result, R limitation of the model was that no information on the variation of the MB shell thickness was obtained. Still, determining the average MB shell thickness is very useful for selecting the optimal imaging conditions. As shown in simulations using MBs with known shell thickness and radius, the difference between the true and calculated shell thickness was within the range of 1030 nm.
3.5
Other applications for the model
The model described in this study has been used for determining the shell thickness of two types of MBs. Theoretically, this model could also be applied for determining the shell thickness of other types of MBs. Other applications for using this model could be for estimating the cell-wall thicknesses of biological samples or shell thicknesses of microcapsules and other spherical structures. In the emerging field of nanotechnology, parameters of pharmaceutical substances, such as the shell thickness, need to be evaluated before regulatory approval. Since images can be recorded at high magnification using TEM, this model can in principle be used for determining shell thicknesses of such nanometer-sized substances.
11
4. Conclusions This study shows that the average shell thickness of polymer-shelled MBs could be determined from TEM images of cross-sectioned MBs using the mathematical model described herein. The model was validated using simulated slices of MBs with known shell thickness and radius and the calculated average shell thickness correlated with the known values. The average shell thickness of Type A MBs, with SPIONs covalently attached to the MB shell surface, was 651 nm. Whereas the average shell thickness of Type B MBs, with SPIONs embedded inside the shell, was 637 nm. This model could potentially be used for determining shell thicknesses of microcapsules and nanometer-sized pharmaceutical substances.
Acknowledgements This study was a part of the 3MiCRON (245572) project, which was founded by the European Commission within the Seventh Framework Program. We thank Dr Kjell Hultenby and his staff at the Electron Microscopy Unit (EMil) at the Karolinska University Hospital (Huddinge, Sweden) for preparing the MB thin sections. We also thank Prof. Gaio Paradossi and his colleagues at the Department of Chemical Sciences and Technologies at the Università di Roma Tor Vergata (Rome, Italy) for preparing Type A and Type B MBs. Support from the Center for Innovative Medicine (CIMED), Karolinska Institutet (Huddinge, Sweden) is also acknowledged.
12
References Barrefelt, Å.A., Brismar, T.B., Egri, G., Aspelin, P., Olsson, A., Oddo, L., Margheritelli, S., Caidahl, K., Paradossi, G., Dähne, L., Barrefelt, A., Brismar, T.B., Egri, G., Aspelin, P., Olsson, A., Oddo, L., Margheritelli, S., Caidahl, K., Paradossi, G., Dahne, L., Axelsson, R., Hassan, M., 2013. Multimodality imaging using SPECT/CT and MRI and ligand functionalized 99mTc-labeled magnetic microbubbles. EJNMMI Res. 3, 12. doi:10.1186/2191-219X-3-12 Brismar, T.B., Grishenkov, D., Gustafsson, B., Härmark, J., Barrefelt, A., Kothapalli, S.V.V.N., Margheritelli, S., Oddo, L., Caidahl, K., Hebert, H., Paradossi, G., Harmark, J., Barrefelt, A., Kothapalli, S.V.V.N., Margheritelli, S., Oddo, L., Caidahl, K., Hebert, H., Paradossi, G., 2012. Magnetite nanoparticles can be coupled to microbubbles to support multimodal imaging. Biomacromolecules 13, 1390–1399. doi:10.1021/bm300099f Cavalieri, F., El Hamassi, A., Chiessi, E., Paradossi, G., 2005. Stable polymeric microballoons as multifunctional device for biomedical uses: Synthesis and characterization. Langmuir 21, 8758–8764. doi:10.1021/la050287j Christiansen, C., Kryvi, H., Sontum, P., Skotland, T., 1994. Physical and biochemical characterization of Albunex, a new ultrasound contrast agent consisting of air-filled albumin microspheres suspended in a solution of human albumin. Biotechnol. Appl. Biochem. 19, 307–320. doi:10.1111/j.1470-8744.1994.tb00300.x Cohen, J.L., Cheirif, J., Segar, D.S., Gillam, L.D., Gottdiener, J.S., Hausnerova, E., Bruns, D.E., 1998. Improved left ventricular endocardial border delineation and opacification with Optison (FS069), a new echocardiographic contrast agent: results of a phase III multicenter trial. J. Am. Coll. Cardiol. 32, 746–752. doi:10.1016/S0735-1097(98)00311-8 Correas, J.-M., Bridal, L., Lesavre, A., Méjean, A., Claudon, M., Hélénon, O., 2001. Ultrasound contrast agents: properties, principles of action, tolerance, and artifacts. Eur. Radiol. 11, 1316–1328. doi:10.1007/s003300100940 du Toit, L.C., Govender, T., Pillay, V., Choonara, Y.E., Kodama, T., 2011. Investigating the Effect of Polymeric Approaches on Circulation Time and Physical Properties of Nanobubbles. Pharm. Res. 28, 494–504. doi:10.1007/s11095-010-0247-y Gao, Z., Kennedy, A.M., Christensen, D.A., Rapoport, N.Y., 2008. Drug-loaded nano/microbubbles for combining ultrasonography and targeted chemotherapy. Ultrasonics 48, 260–270. doi:10.1016/j.ultras.2007.11.002 Gramiak, R., Shah, P.M., 1968. Echocardiography of the aortic root. Invest. Radiol. 3, 356–366. Hansson, A., Hance, N., Dufour, E., Rantanen, A., Hultenby, K., Clayton, D.A., Wibom, R., Larsson, N.-G.G., 2004. A switch in metabolism precedes increased mitochondrial biogenesis in respiratory chain-deficient mouse hearts. Proc. Natl. Acad. Sci. U. S. A. 101, 3136–3141. doi:10.1073/pnas.0308710100 Harris, J.R., Depoix, F., Urich, K., 1995. The structure of gas-filled n-butyl-2-cyanoacrylate 13
(BCA) polymer particles. Micron 26, 103–111. doi:10.1016/0968-4328(95)93275-H Härmark, J., Larsson, M.K., Razuvajev, A., Koeck, P.J.B., Paradossi, G., Brodin, L.-Å., Caidahl, K., Hebert, H., Bjällmark, A., 2015. Investigation of the elimination process of a multimodal polymer-shelled contrast agent in rats using ultrasound and transmission electron microscopy. Biomed. Spectrosc. Imaging 4, 81–93. doi:10.3233/BSI-140099 Kovalenko, A., Jouhannaud, J., Polavarapu, P., Krafft, M.P., Waton, G., Pourroy, G., 2014. Hollow magnetic microspheres obtained by nanoparticle adsorption on surfactant stabilized microbubbles. Soft Matter 10, 5147–5156. doi:10.1039/c4sm00318g Mai, Y., Eisenberg, A., 2010. Controlled Incorporation of Particles into the Central Portion of Vesicle Walls. J. Am. Chem. Soc. 132, 10078–10084. doi:10.1021/ja1024063 Mercadé-Prieto, R., Nguyen, B., Allen, R., York, D., Preece, J.A., Goodwin, T.E., Zhang, Z., 2011. Determination of the elastic properties of single microcapsules using micromanipulation and finite element modeling. Chem. Eng. Sci. 66, 2042–2049. doi:10.1016/j.ces.2011.01.015 Mikhaylova, M., Kim, D.K., Berry, C.C., Zagorodni, A., Toprak, M., Curtis, A.S.G., Muhammed, M., 2004. BSA immobilization on amine-functionalized superparamagnetic iron oxide nanoparticles. Chem. Mater. 16, 2344–2354. doi:10.1021/cm0348904 Morel, D.R., Schwieger, I., Hohn, L., Terrettaz, J., Llull, J.B., Cornioley, Y.A., Schneider, M., 2000. Human Pharmacokinetics and Safety Evaluation of SonoVueTM, a New Contrast Agent for Ultrasound Imaging. Invest. Radiol. 35, 80–5. doi:10.1097/00004424-20000100000009 Narayan, P., Wheatley, M.A., 1999. Preparation and characterization of hollow microcapsules for use as ultrasound contrast agents. Polym. Eng. Sci. 39, 2242–2255. doi:10.1002/pen.11612 Nguyen, P.N., Nikolova, G., Polavarapu, P., Waton, G., Phuoc, L.T., Pourroy, G., Krafft, M.P., 2013. Compressible multi-scale magnetic constructs: decorating the outer surface of selfassembled microbubbles with iron oxide nanoparticles. RSC Adv. 3, 7743–7746. doi:10.1039/c3ra40908b Poehlmann, M., Grishenkov, D., Kothapalli, S.V.V.N., Härmark, J., Hebert, H., Philipp, A., Hoeller, R., Seuss, M., Kuttner, C., Margheritelli, S., Paradossi, G., Fery, A., 2014. On the interplay of shell structure with low- and high-frequency mechanics of multifunctional magnetic microbubbles. Soft Matter 10, 214–226. doi:10.1039/C3SM51560E Sanna, V., Pintus, G., Bandiera, P., Anedda, R., Punzoni, S., Sanna, B., Migaleddu, V., Uzzau, S., Sechi, M., 2011. Development of Polymeric Microbubbles Targeted to Prostate-Specific Membrane Antigen as Prototype of Novel Ultrasound Contrast Agents. Mol. Pharm. 8, 748– 757. doi:10.1021/mp100360g Schneider, M., 1999. SonoVue, a new ultrasound contrast agent. Eur. Radiol. 9, S347–S348. doi:10.1007/pl00014071 Schneider, M., Arditi, M., Barrau, M.-B., Brochot, J., Broillet, A., Ventrone, R., Yan, F., 1995. 14
BR1: a new ultrasonographic contrast agent based on sulfur hexafluoride-filled microbubbles. Invest. Radiol. 30, 451–457. Smith, A.E., Zhang, Z., Thomas, C.R., 2000. Wall material properties of yeast cells: Part 1. Cell measurements and compression experiments. Chem. Eng. Sci. 55, 2031–2041. doi:10.1016/S0009-2509(99)00500-X Villa, R., Cerroni, B., Viganò, L., Margheritelli, S., Abolafio, G., Oddo, L., Paradossi, G., Zaffaroni, N., 2013. Targeted doxorubicin delivery by chitosan-galactosylated modified polymer microbubbles to hepatocarcinoma cells. Colloids Surfaces B Biointerfaces 110, 434–442. doi:10.1016/j.colsurfb.2013.04.022 Vlaskou, D., Mykhaylyk, O., Krötz, F., Hellwig, N., Renner, R., Schillinger, U., Gleich, B., Heidsieck, A., Schmitz, G., Hensel, K., Plank, C., 2010. Magnetic and Acoustically Active Lipospheres for Magnetically Targeted Nucleic Acid Delivery. Adv. Funct. Mater. 20, 3881–3894. doi:10.1002/adfm.200902388 Yang, F., Li, Y.X., Chen, Z.P., Zhang, Y., Wu, J.R., Gu, N., 2009. Superparamagnetic iron oxide nanoparticle-embedded encapsulated microbubbles as dual contrast agents of magnetic resonance and ultrasound imaging. Biomaterials 30, 3882–3890. doi:10.1016/j.biomaterials.2009.03.051 Zimny, K., Mascaro, B., Brunet, T., Poncelet, O., Aristégui, C., Leng, J., Sandre, O., MondainMonval, O., 2014. Design of a fluorinated magneto-responsive material with tuneable ultrasound scattering properties. J. Mater. Chem. B 2, 1285–1297. doi:10.1039/c3tb21585g
15
Figure 1 Transmission electron microscopy images of a Type A (A) and Type B (B) microbubbles. The dark particles (arrows) are superparamagnetic iron oxide nanoparticles attached to (A), or embedded in (B), the polymer shell. Scale bars represent 500 nm.
Figure 2 A theoretical section through the epoxy resin (grey). As illustrated, slicing at varying distances from the MB center will affect the observed shell thickness (T1 > T2).
Figure 3 Illustrating the parameters used in our model for determining the shell thickness of spherical objects. T is defined as the shell thickness, R as the outer radius, r as the inner radius, Ro as the observed outer radius, ro as the observed inner radius, z as the projected y-axis height at slicing angles φ and θ and, finally To as the observed shell thickness.
16
Fig.1
17
Fig.2
18
Fig.3
19
̅ of two types of microbubbles determined by the model Table 1 Average shell thickness 𝐓
Microbubbles
̅ 𝐨 [nm] 𝐑
𝐫̅𝐨 [nm]
̅ [nm] 𝐑
𝐫̅ [nm]
̅ [nm] 𝐓
Type A
1478
967
1882
1231
651
Type B
1421
920
1809
1171
637
̅) and inner (r̅) radius were calculated from the mean observed outer (R ̅ o ) (n=100 for Type A and Type B) Notes: the mean outer (R and inner (r̅o) (n=80 for Type A and n=82 for Type B) radius of the microbubbles in the transmission electron microscopy sections.
20
Table 2 The average shell thickness and standard deviation calculated from the model by iteratively (n = 10) simulate cutting random slices (n = 1000) of microbubbles with known shell thickness and radius.
Shell thickness [nm] Radius [nm]
200
400
600
800
1000
1500
204±10
394±11
601±16
800±14
1000±15
2000
204±10
404±13
602±15
795±15
996±17
2500
204±14
405±17
607±27
800±24
993±34
21