2.1.

Materials

In our experiments, SWNTs were purchased from the Chinese Academy of Sciences (Chengdu, China). Nude mice were purchased from Sun Yat-Sen University (Guangzhou, China). Phospholipid- $\text{Poly}\left(\text{ethylene}\right){\text{glycol}}_{2000}$ (PL-PEG) was purchased from Avanti Polar Lipids, Inc. (Shanghai, China). Integrin ${\alpha }_{\mathrm{v}}{\beta }_3$ antibody was purchased from Santa Cruz Biotechnology Inc. (Santa Cruz, California, USA), and Protein A was purchased from Bei Jing Biosynthesis Biotechnology Co., Ltd. (Beijing, China).

2.2.

Administration of SWNTs

SWNTs were sonicated in an aqueous solution of PL-PEG at a ratio of $1\mathrm{mg}$ SWNTs: $1\mathrm{mg}$ PL-PEG: $1\mathrm{mL}$ water for $6\mathrm{h}$ , followed by centrifugation at $1000\mathrm{g}$ for $1\mathrm{h}$ . Then, the supernatant was collected and underwent an additional centrifugation at $\mathrm{24,000}\mathrm{g}$ for $6\mathrm{h}$ to obtain well-suspended PL-PEG-functionalized SWNTs in the supernatant. Excess phospholipids were thoroughly removed by repeated filtration through $100\mathrm{kDa}$ filters (Millipore) and rinsed thoroughly with water, then resuspended SWNTs in either water or phosphate-buffered saline (PBS) by sonication for $1\mathrm{h}$ . The PL-PEG-functionalized SWNTs (denoted as SWNT-PEG-COOH) were finally resuspended in PBS.

2.3.

Anti-integrin ${\alpha }_{\mathrm{v}}{\beta }_3$ Conjugation to SWNTs

SWNT-PEG-COOH solution was reacted with $N$ -hydroxysulphonosuccinimide (NHS) in the presence of 1-ethyl-3-[3-(dimethylamino)-propyl] carbodiimide at 1:1:1 molar ratio, then Protein A was mixed with the SWNT-PEG-COOH solution at pH 7.4 and incubated for $4\mathrm{h}$ . The SWNT-PEG-Protein A was filtrated through $100\mathrm{KDa}$ filters (Millipore) to remove excess Protein A. Then, integrin ${\alpha }_{\mathrm{v}}{\beta }_3$ antibody was added into a solution of SWNT-PEG-Protein A. The mixture was allowed to react for $4\mathrm{h}$ at room temperature and protected from light. After reaction, the solution was dialyzed against PBS using a membrane (molecular weight $\text{cutoff}=500\mathrm{KDa}$ ) to remove dissociative antiintegrin ${\alpha }_{\mathrm{v}}{\beta }_3$ . The dialysis was carried out for $3\text{days}$ with frequent replacement of the buffer to obtain the final sample SWNT-PEG-Anti-integrin ${\alpha }_{\mathrm{v}}{\beta }_3$ solution (see Fig. 1 ).

Fig. 1

Water-soluble SWNT functionalized with PEG and integrin ${\alpha }_{\mathrm{v}}{\beta }_3$ antibodies. Schematic drawings of noncovalently functionalized SWNT- ${\mathrm{PEG}}_{2000}$ with Protein A and integrin ${\alpha }_{\mathrm{v}}{\beta }_3$ antibody. The hydrophobic carbon chains of the phospholipids strongly bind to the sidewalls of the SWNTs, and the PEG chains render water solubility to the SWNTs. The integrin ${\alpha }_{\mathrm{v}}{\beta }_3$ antibody on the SWNTs are used to specifically target the integrin ${\alpha }_{\mathrm{v}}{\beta }_3$ -positive tumor.

2.4.

Cell Lines and the Animal Model

U87MG human glioblastoma cancer cell lines (from American Type Culture Collection) were cultured under standard conditions. The U87MG tumor models were generated by subcutaneous injection of $5\times {10}^6$ cells in $50\mu \mathrm{L}$ PBS into the back of the nude balb/c mice. PEGylated SWNTs were intramuscular-injected into the tumor-free nude mice tail for the in vivo PA molecular imaging experiment. Targeted SWNTs with anti-integrin ${\alpha }_{\mathrm{v}}{\beta }_3$ molecules were intravenously injected into the tumor-bearing nude mice tail veins for tumor-targeting experiment

2.5.

PA Molecular Imaging System for In Vivo SWNTs Selective Targeting

The PA molecular imaging system for in vivo SWNTs selective targeting is shown in Fig. 2 . The photoacoustic signal is captured by a linear ultrasound transducer array (L7L38A, SIUI). The 128 elements piezoelectric linear transducer has a $7.5\text{-}\mathrm{MHz}$ central frequency and a 70% nominal bandwidth. The physical dimensions of the elements are given by an elemental elevation aperture of $10\mathrm{mm}$ and a lateral pitch of $300\mu \mathrm{m}$ . Each element of the transducer array has a thin cylinder ultrasonic lens that produces a geometric focus approximately $35\mathrm{mm}$ in front of the transducer array to select the 2-D image plane and suppress the out-of-plane signals.

Fig. 2

Schematic of the experimental setup for the photoacoustic molecular imaging system. For optical absorption distribution reconstruction, the photoacoustic signal is captured by a linear ultrasound transducer array (L7L38A, SIUI). The 128 elements piezoelectric linear transducer has a $7.5\text{-}\mathrm{MHz}$ central frequency and a 70% nominal bandwidth. An OPO-provided $750\text{-}\mathrm{nm}$ laser beam is expanded and homogenized by the concave lens and the ground glass to irradiate on the backs of the mice. The linear ultrasound transducer array, driven by a computer-controlled step motor to scan around the mice, captures the PA signal after targeted SWNTs injection.

A frequency-doubled Nd:yttrium aluminum garnet laser pumping an optical parametric oscillator (Vibrant 532 I, Opotek, Carlsbad, CA) was employed to provide $690–960\mathrm{nm}$ laser light with a pulse duration of $10\mathrm{ns}$ and a pulse repetition rate of $10\mathrm{Hz}$ . The laser beam was expanded and homogenized to provide an incident energy density of $<10\mathrm{mJ}∕{\mathrm{cm}}^2$ and a beam diameter of $1\mathrm{cm}$ to irradiate the sample for generating photoacoustic signals. A $25\mathrm{MHz}$ clock signal, provided by a control circuit, was divided into $10\mathrm{Hz}$ for triggering the pulse laser and controlling the multielement linear transducer array to acquire photoacoustic signals. The signals from the transducers, after preprocessing, including ultralow-noise time-gain compensate amplifiers, anti-aliasing bandpass filters, analog-to-digital converters, and several beamforming strategies (various f-number, aperture apodization, element size, etc.), were acquired with the NI PCI-6541 High-Speed Digital Wfm ( $50\mathrm{MHz}$ , selectable; voltage, 32 channels, $1\mathrm{Mbit}$ /channel). The collected data were stored in a computer for later processing. Next, the multielement linear transducer array was driven by a step motor to circularly scan around the sample, detecting the photoacoustic signals in the imaging plane at each position. The linear transducer array and the samples were both immersed in a tank of water for better coupling. The induced photoacoustic waves were captured every $18\mathrm{deg.}$ , and a total of 20 positions of photoacoustic waves are recorded for a full view of the circularly scanning angle. After the $128\times 20$ series data were transferred to the computer, further postprocessing was done using the MATLAB program (version 7.0, Mathworks). Projections were calculated with the improved limited-field filtered back projection algorithm.¹⁶ At last, the photoacoustic images were shown on the display.

3.1.

Optical Properties of SWNTs for PA Molecular Imaging

SWNTs were dispersed in the aqueous phase by noncovalently adsorbing PL-PEGs. Figure 3 presents visible spectrometry (VIS)/NIR absorption spectra of the initial blood and functionalized SWNT samples in water. Absorption spectra of functionalized SWNTs were compared with pure blood absorption at the same concentration. The two absorption spectra are significantly different and clearly show that the SWNT can enhance absorption in the NIR region, in which the biological tissue is almost transparent with very low absorbance. Longer wavelengths are more desirable, as the depth of penetration through the tissues is increased; the photoacoustic spectra suggest that $750\mathrm{nm}$ is the preferable wavelength, because the photoacoustic signal of the SWNTs is higher than blood absorption at that wavelength. The high absorbance of SWNTs in the NIR originates from electronic transitions between the first or second van Hove singularities of the nanotubes.¹⁷ High optical absorbance of SWNTs in the $700–1100\text{-}\mathrm{nm}$ NIR window transparent to biological systems is exploited in the current work at a single wavelength by using a $750\text{-}\mathrm{nm}$ pulse laser for radiation.

Fig. 3

VIS/NIR absorption spectra of the initial blood and functionalized SWNT samples in water. Absorption spectra of functionalized SWNTs were compared with pure blood absorption at the same concentration. The two absorption spectra are significantly different and clearly show that the SWNT can enhance absorption in the NIR region, where the optical transmission through biological tissues is optimal.

In contrast to NIR-absorbing dyes, the absorption properties of SWNTs are dependent upon a rigid structure rather than on molecular orbital electronic transitions. SWNTs are also not susceptible to photobleaching, a problem commonly associated with other absorbing dyes. These features make SWNTs suitable as a contrast agent for photoacoustic molecular imaging in the NIR region.

3.2.

SWNTs for Photoacoustic Molecular Imaging

This experiment was performed to determine the concentration of SWNTs as a NIR contrast agent for effective enhancement in photoacoustic molecular imaging. The optical absorption spectrum of PL-PEG functionalized SWNT was tested at different concentrations [see Fig. 4a ]. Linear curve fitting for peak absorption at a wavelength of $750\mathrm{nm}$ was shown in Fig. 4b. For the PA signal peak intensity of different concentrations of SWNT solution, linear curve fitting for $750\mathrm{nm}$ absorption was evaluated to estimate the concentration of SWNTs [see Fig. 4c]. Different concentrations of SWNT solution mixed with blood were also investigated in vitro to estimate the appropriate concentration for in vivo application in Fig. 4d.

Fig. 4

SWNTs as IR contrast agent with varying concentrations for photoacoustic molecular imaging. (a) Absorption spectra of ${\mathrm{PEG}}_{2000}$ -functionalized SWNT suspensions in PBS with varying concentrations. (b) Linear fitting of peak absorbance of spectrum versus concentration of SWNTs at $750\mathrm{nm}$ . (c) PA signal peak intensity of different concentrations of SWNT solution; linear curve fitting for absorption of $750\mathrm{nm}$ was evaluated to estimate the concentration of SWNTs. (d) Different concentrations of SWNT solution suspended in blood in vitro. The series of experiments indicated that a SWNT solution concentration of $0.17\mathrm{mg}∕\mathrm{mL}$ can achieve effective enhancement.

These data suggest that a SWNT solution concentration of 0. $17\mathrm{mg}∕\mathrm{mL}$ can achieve effective enhancement using a $750\mathrm{nm}$ pulse laser. We thus expect, based on the concentrations of SWNTs expected to be bound in vivo, that it will be possible to detect the uptake of the proposed in vivo tumor probe.

3.3.

In Vitro PA Measurements

SWNTs as IR contrast agent for photoacoustic molecular imaging were demonstrated in an in vitro experiment. Two identical $0.8\text{-}\mathrm{mm}$ -diameter tubes, one with $300\mu \mathrm{L}$ SWNTs of $0.17\mathrm{mg}∕\mathrm{mL}$ and one with pure blood, were placed side by side in the imaging tank at a depth of $10\mathrm{mm}$ in a mimic phantom of Intralipid-20% dilution. The image in Fig. 5a was obtained by photoacoustic scanning by the linear transducer array and represents a clearer map of tube A containing $\mathrm{SWNTs}+\text{blood}$ than of tube B containing blood only. With the exogenous contrast agent, the optical absorption of the blood was increased, and the contrast between the vessels and the background brain tissues was enhanced. Reconstructed profiles of these two tubes shown in Fig. 5b also show higher signal-to-noise ratios using a SWNT-enhanced sample than using blood only at the excitation wavelength of $750\mathrm{nm}$ .

Fig. 5

SWNTs as IR contrast agent for photoacoustic imaging in vitro. (a) Photoacoustic images of A gel phantom with two tubes, tube A containing $\mathrm{SWNTs}+\text{blood}$ and tube B containing blood only. (b) The line normal to object axis profile of the reconstructed image shown in (a) with $x=0\mathrm{mm}$ , which shows a better imaging contrast with enhanced SWNTs than blood only at a wavelength of $750\mathrm{nm}$ .

3.4.

Cellular Toxicity Tests of SWNTs

Cell cytotoxicity assay was performed with a colorimetric tetrazolium salt-based assay, Cell Counting Kit-8 (CCK8). To determine the cytotoxicity of SWNTs, tumor cells ( $1\times {10}^3$ per well) were cultured in a 96-well microplate for $24\mathrm{h}$ and then co-incubated with functionalized SWNTs of different concentrations for $12\mathrm{h}$ , rinsed with PBS, and incubated for another $72\mathrm{h}$ . OD450, with an absorbance value of $450\mathrm{nm}$ , was read with a 96-well plate reader (Infinite M200, Tecan, Switzerland) to determine the viability of the cells. The viability of cells was calculated as: cell viability $(\%\text{of}\text{control})=\mathrm{ODTre}∕\mathrm{ODCon}\times 100\%$ (where ODTre was the absorbance value at $450\mathrm{nm}$ of treated cells and ODCon was the absorbance value at $450\mathrm{nm}$ of control cells). Cell-toxicity tests (see Fig. 6 ) show that SWNTs display satisfactory biosafety at the dosages expected for in vivo applications.

Fig. 6

Cell toxicity of SWNT measurement. The viability of cells was calculated as: cell viability $(\%\text{of}\text{control})=\mathrm{ODTre}∕\mathrm{ODCon}\times 100\%$ (where ODTre was the absorbance value at $450\mathrm{nm}$ of treated cells and ODCon was the absorbance value at $450\mathrm{nm}$ of control cells). SWNT-treated cells show little decreased viability in the colorimetric tetrazolium salt-based assay at $4\text{days}$ when compared to control cells (ODCon). With the CCK8 assay, at all concentrations, SWNT-treated cells also show no significantly decreased viability when compared to control cells.

3.5.

In Vivo PA Imaging of SWNTs

SWNTs as IR contrast agent for photoacoustic molecular imaging were demonstrated in an in vivo experiment. Female nude mice ( $20–25\mathrm{g}$ body weight) were employed in this work. General anesthesia was administered on the mice by an intramuscular injection of ketamine hydrochloride $(44\mathrm{mg}∕\mathrm{kg})$ , xylazine hydrochloride $(2.5\mathrm{mg}∕\mathrm{kg})$ , acepromazine maleate $(0.75\mathrm{mg}∕\mathrm{kg})$ , and atropine $(0.025\mathrm{mg}∕\mathrm{kg})$ . During the experiment, the mouse was placed on a water-circulating heating pad, and additional heating was provided by an overhead surgical lamp. During the data acquisition, the rat was provided pure oxygen for breathing, and the arterial blood oxygenation $\left(\mathrm{Sp}{\mathrm{O}}_2\right)$ level and heart rate were monitored by a pulse oximeter ( $8600\mathrm{V}$ , Nonin) with the fiberoptic probe wrapped around a paw of the animal. The SWNT was successively intramuscular-injected into the tail of the mouse at a dose of $300\mu \mathrm{L}$ SWNTs dispersed in PBS with a concentration of $0.17\mathrm{mg}∕\mathrm{mL}$ . Image acquisition of the mouse tail began approximately $120\min$ following administration.

Two of the photoacoustic angiographs of the rat tail are presented in Fig. 7a and 7c . Compared to the tail image based on the intrinsic optical contrast [Fig. 7a], the image with SWNTs contrast agent [Fig. 7c] shows greater clarity. It also shows a better signal-to-noise ratio when comparing Fig. 7b with 7d.

Fig. 7

In vivo PA imaging of SWNTs as contrast agent in mouse tail. (a) Control PA image without SWNTs- ${\mathrm{PEG}}_{2000}$ injection. (b) Reconstruction profile of mouse tail shown in (a). (c) PA image at $2\mathrm{h}$ after SWNTs-PEG injection. (d) Reconstruction profile of mouse tail shown in Fig. 7d.

3.6.

In Vivo PA Imaging of SWNTs Targeted Tumor

In vivo PA imaging of SWNTs targeted tumor has been performed in this experiment. The targeted SWNTs with anti-integrin ${\alpha }_{\mathrm{v}}{\beta }_3$ molecules and SWNTs-PEG (control group) were injected into the tail veins of two mice. In each mouse, the targeting process was imaged $2\mathrm{h}$ after injection.

Photoacoustic scanning was performed after tumor-cell inculcation for two weeks. The PA images of a U87 tumor for injections with targeted SWNTs and SWNTs-PEG (control group) in Fig. 8 demonstrate the specific targeting ability of the targeted SWNTs probe. The contrast is higher for the targeted SWNTs data postinjection ( $2\mathrm{h}$ after injection) within the tumor region [Fig. 8b] than for the SWNTs-PEG injection [Fig. 8a]. The contrast between the experimental group (targeted SWNTs injection) and the control group (SWNTs-PEG injection) indicated specific targeting of SWNTs targeted with anti- ${\alpha }_{\mathrm{v}}{\beta }_3$ to U87 cells.

Fig. 8

In vivo PA imaging of U87 tumor before and after injection of targeted SWNTs. The PA signals were generated by pulsed laser at a wavelength of $750\mathrm{nm}$ . (a) Photograph of U87 tumor on the back of nude mice. Dotted line shows the tumor area. (b) PA image at $2\mathrm{h}$ after SWNTs-PEG injection. (c) PA image at $2\mathrm{h}$ after SWNT-PEG2000-Protein A-anti-integrin $\alpha \mathrm{v}{\beta }_3$ injection.