Experimental study of ultrasound retention of bubble-surrounded cells under various conditions of acoustic field and flow velocity

In this experiment, we used CD8 positive T lymphocytes (the cells, hereinafter), ⁹⁾ which recognize an antigen and become activated to kill infected or malignant cells. The cells have a size range around 10–12 μm and were dyed with tetramethyl rhodamine. Also, we prepared anti-CD8 antibody-modified bubble liposomes (bubbles, hereinafter), which have the antibody to link covalently on the surface of the cells. As well as in our preceding research, ^12,27) we prepared a suspension of BSCs, diluted in phosphate-buffered saline (PBS) to the desired concentrations of cells, and the bubbles, and stirred using a glass rod in a beaker for up to 60 s. To determine the concentration of the cells quantitatively, we sampled the suspension of 10 μl and observed the contained number of the cells using C-CHIP (NanoEntek).

In this research, first, we tried to estimate the concentration of the cells using the fluorescence intensity derived from the brightness in a microscopic image. Although estimating the number of layered cells in the perpendicular direction is difficult using a conventional optical observation, there is a possibility that the fluorescent magnitude increases in proportion to the number of layered cells. We prepared an artificial blood vessel made of polycarbonate covered by a polydimethylsiloxane sheet, which has an acoustic impedance of 9.8 × 10⁵ kg (m⁻² s⁻¹) and three paths with each thickness of 1.0 mm, as shown in Fig. 1. We chose the straight path in the center, which has the width α = 1 mm and the length β = 80 mm, and filled it with the suspension to observe the fluorescent images using an industrial microscope (Olympus BXFM) with a digital camera (Olympus DP74).

Zoom In Zoom Out Reset image size

Here, defining the concentration of the cells as C and the volume of the path as V, the area density of the cells contained in the path was calculated as CV/αβ. The number of the cells was obtained by integrating the above area density in the whole path. Considering a x–y coordinate in the obtained fluorescent image, the number of the cells N in the path should be

$$\begin{eqnarray}&&N=\displaystyle \frac{V\gamma }{\alpha \beta }\,\displaystyle \sum _{i}\displaystyle \sum _{j}{C}_{ij},\end{eqnarray} \tag{ 1 }$$

where C_ij indicates the concentration of the cells in the pixel (i, j), and γ indicates the area of a pixel. In this equation, because the parameters of N, V, γ, α, and β are known by the examiner, if the path is filled with a suspension and the brightness on the fluorescent image is regarded to be uniform, then the relationship between mean brightness B and mean the concentration 〈C_ij 〉 should be derived, where 〈 〉 indicates the average in the observation area.

We have designed the experimental setup ^11,12) to observe the behavior of the BSCs in flow under ultrasound exposure. Figure 2 shows the vertical view of the experimental setup including the same fluorescence microscope and digital camera, a two-dimensional array transducer, a water tank, and the artificial blood vessel. The transducer, which has a central frequency of 3 MHz and 128 elements, ²⁸⁾ was installed at the bottom of the water tank and targeted the observation area with a distance of l = 60 mm. The elevation angle θ was set to 60° such that the irradiation area of the acoustic field was included in the observation area. The artificial blood vessel was placed at the water surface. In the following experiments, the concentrations of the cells and bubbles were fixed to be 1.0 × 10⁵ ml⁻¹ and 0.3 mg ml⁻¹ in the suspension, respectively, a condition in which we confirmed, under the microscopic images using a confocal microscope, that a cell was fully covered by bubbles. ³⁰⁾ Under the artificial flow of PBS through the artificial blood vessel, a 0.5 ml suspension of BSCs was injected at a point 350 mm upstream from the observation area during ultrasound irradiation, where a duration of ultrasound emission was fixed at 30 s. We chose the velocity range 10–30 mm s⁻¹ because the flow velocity near the main hepatic artery was assumed to be approximately 100–200 mm s⁻¹. Then the cloudy adhesion of BSCs on the upper wall of the path was observed in the middle of the observation area to evaluate the retention performance of BSCs. The number of the retained cells was obtained by a conversion from the brightness of acquired images using Eq. (1) and image analysis software (Nippon Roper, Image pro plus).

Zoom In Zoom Out Reset image size

To investigate the method of ultrasound emission to retain as many cells as possible, we applied the tempo-spatial division emission, ²⁸⁾ which moves the position of a single focal point electrically by changing the delay time, in the observation area. Figure 3 shows the positions of the focal points, which move repeatedly from f₁ to f_n , where the value n indicates the number of focal points. Because the driving equipment (Microsonic ES1144-1) generates a burst wave with a maximum duty ratio of 60%, pulse repetition time was fixed at 1.7 ms, where a continuous wave was emitted for 1.0 ms in each period. For when the number n is more than 2, the transition rate v [mm s⁻¹] between the focal points, which is defined by dividing the distance r [mm] by t_e , where t_e indicates the emission duration in each focal point.

Zoom In Zoom Out Reset image size

As mentioned in the previous section, the distribution of acoustic intensity on the path varies according to the number of focal points and the magnitude of sound pressure. Therefore, to evaluate retained cells quantitatively, we defined and calculated the applied acoustic intensity of tempo-spatial division emission by measuring actual acoustic fields. Figure 4 shows an example of an acoustic field of a single focal point, where the distribution was divided into small grid areas, with a width w. Also, defining mean sound pressure as P_ij [Pa-pp] in a small area, S_ij , the applied acoustic power E can be calculated by

$$\begin{eqnarray}&&\begin{array}{l}E=\displaystyle \sum _{i}\displaystyle \sum _{j}\displaystyle \frac{0.6}{n}\,{w}^{2}\displaystyle \frac{1}{2{\rm{\pi }}}\displaystyle {\int }_{0}^{2\pi }\displaystyle \frac{{\left(\displaystyle \frac{{P}_{ij}}{2}\,\sin \,\theta \right)}^{2}}{Z}d\theta \\ \,=\,\displaystyle \frac{0.6{w}^{2}}{8nZ}\displaystyle \sum _{i}\displaystyle \sum _{j}{P}_{ij}^{2},\end{array}\end{eqnarray} \tag{ 2 }$$

where Z is an acoustic impedance of the medium. The coefficient 0.6/n was chosen because the duty ratio per focal point was limited to 60% for the equipment, as mentioned in the previous section.

Zoom In Zoom Out Reset image size

Generally, since acoustic power increases in proportion to the exposed area, it is necessary to normalize it by the area, as well as spatial average temporal average intensity, in order to compare between various acoustic fields. In addition to the evaluation of thermal effects caused by high-intensity focused ultrasound, we defined the applied acoustic intensity as I = E/S, where S indicates the area of the acoustic field, which is defined by the area where the relative amplitude is more than −20 dB of the focal point. ³¹⁾ For the calculation of both E and S, an identical coordinate of w = 0.5 mm was used.

As mentioned in Sect. 2.1, we prepared multiple suspensions, with various concentrations of the cells ranging from 1.0 × 10⁵ ml⁻¹ to 90.0 × 10⁵ ml⁻¹, and concentration of bubbles set to 0.3 mg ml⁻¹, to investigate the relationship between the mean brightness of the fluorescent image and the concentration. Figure 5(a) shows three fluorescent images of the suspension-filled paths without flow, where the concentrations are 100, 400, and 800 × 10⁴ ml⁻¹ under the microscope settings of an 8x gain and 80 ms exposure time. A Plan-Apochromat 1.25x objective lens (NA 0.04, Olympus) was equipped with the microscope. The mean brightness was increased in proportion to the concentration of the cells. Figure 5(b) shows the relationship between the concentration 〈C_ij 〉 and brightness B, where the brightness saturated when 〈C_ij 〉 was more than 800 × 10⁴ ml⁻¹. From this result, we derived the correlation equation by fitting a quadratic function between 〈C_ij 〉 and B. Here it should be mentioned that the brightness threshold was set at 20 because of the background magnitude, which means that concentrations less than 38.8 × 10⁴ ml⁻¹ were ignored for the cell counting using Eq. (1).

Zoom In Zoom Out Reset image size

Before applying tempo-spatial division emission of a focal point, as mentioned in Sect. 2.3, we measured the sound pressure distributions of various acoustic fields with multiple focal points using an acoustic intensity measurement system (Onda AIMS III) by translating the hydrophone (Onda HNR-1000) in degassed water. Figure 6 shows the profiles of the applied acoustic intensity when the number of focal points was n = 1, 2, 3, and 4, with spatial interval (a) r = 2 mm and (b) r = 3 mm, along the y-axis of the path as shown in Fig. 3. Figure 7 shows the special distribution of the applied acoustic intensity in x–y plane. These distributions were measured individually to superimpose the identical coordinates. Note that although the maximum sound pressure of each focal point was adjusted to 400 kPa-pp, the peak value of the applied acoustic intensity varied according to the number of the focal points because of tempo-spatial division emission.

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Before verifying the parameters of the acoustic field shape, we investigated the effect of the transition rate of the focal points produced by the previously mentioned tempo-spatial division emission. By fixing the ultrasound emission conditions of the three focal points (i.e. spatial interval r of 3 mm, maximum sound pressure of 400 kPa-pp realized with a two-dimensional array transducer, and flow velocity of 10 mm s⁻¹), we examined to retain the BSCs at different transition rates from 10 to 1000 mm s⁻¹. Figure 8 shows the number of retained T-cells versus the transition rate; the error bars indicate the standard deviations. The number of retained cells was maximal when the transition rate was 100 mm s⁻¹; thus, a transition rate that is 50 times faster than the flow velocity does not effectively propel the cells. Therefore, we fixed the transition rate at v = 100 mm s⁻¹ in the following experiments to determine which other parameters can retain cells. Under this condition, because the emission duration t_e in each focal point was 30 ms, the repetition time for the three focal points was 90 ms. As the total duration of the ultrasound emission process was 30 s (Sect. 2.2), the transition of the focal points was repeated more than 300 times.

Zoom In Zoom Out Reset image size

As mentioned in the above section, the distribution of sound pressure depends on the spatial interval of the focal points. Here, we compare the retained number of the cells with two different spatial intervals with multiple focal points using tempo-spatial division emission. The acoustic field was produced with three focal points and a maximum sound pressure of 400 kPa-pp, which are shown in Fig. 6. Figure 9 shows the fluorescent images of the retained BSCs with a flow velocity of 10 mm s⁻¹ taken 30 s after the injection of the suspension. In both spatial intervals, although the cells were not confirmed near the first focal point f₁, clouds of the retained cells were seen near f₂ and f₃. However, when the spatial interval was set to 3 mm, there seemed to be lack of retained cells between f₂ and f₃, which corresponds to the distribution of sound pressure shown in Fig. 9.

Zoom In Zoom Out Reset image size

Figure 10 shows the calculated number of retained cells versus the number of focal points with a maximum sound pressure of 300 kPa-pp according to flow velocities. The error bars indicate the standard deviation. The number of retained cells showed the maximum when two or three focal points were applied with a spatial interval of r = 2 mm and a flow velocity of 10 mm s⁻¹. Also, it is obvious that the number of retained cells significantly decreased with spatial interval r = 3 mm with every number of focal points. Considering the above results, we fixed the spatial interval r = 2 mm in the following experiments to verify other parameters to retain more cells.

Zoom In Zoom Out Reset image size

We investigated similar retention performances by varying the flow velocity, maximum sound pressure, and number of focal points in tempo-spatial division emission. the retained number of the cells increased in proportion to sound pressure. Figure 11 shows the number of retained cells versus the applied acoustic intensity according to the maximum sound pressure, number of focal points and flow velocity. Regarding Figs. 11(a) to 11(c), with flow velocities of (a) 10 mm s⁻¹, (b) 20 mm s⁻¹ and (c) 30 mm s⁻¹, the retained number of the cells increased in proportion to applied acoustic intensity. Comparing the results when the maximum sound pressure was 400 kPa-pp, in Fig. 11(a), a maximum number of 10 000 cells was retained with three focal points, where 80% of the ultrasound-exposed cells was retained in the path. On the other hand, in Fig. 11(b), 6000 cells were retained with two focal points, despite the acoustic condition in the horizontal axis in Fig. 11 being common in all results. Then the maximum retention performance was confirmed with an applied acoustic intensity of 40–80 mW cm⁻² with a flow velocity of 10 mm s⁻¹, whereas the applied acoustic intensity of 120 mW cm⁻² was the maximum with a flow velocity of 20 mm s⁻¹. Also, the number of retained cells varied with applied acoustic intensity in higher number of focal points, where the multiple focal points with an acoustic intensity of 50–120 mW cm⁻² were more effective than the single focal point with 180 mW cm⁻² in the range of a flow velocity of 10–20 mm s⁻¹.

Zoom In Zoom Out Reset image size