Golden carbon nanotubes as multimodal photoacoustic and photothermal high-contrast molecular agents

Abstract

Carbon nanotubes have shown promise as contrast agents for photoacoustic and photothermal imaging of tumours and infections because they offer high resolution and allow deep tissue imaging. However, in vivo applications have been limited by the relatively low absorption displayed by nanotubes at near-infrared wavelengths and concerns over toxicity. Here, we show that gold-plated carbon nanotubes-termed golden carbon nanotubes-can be used as photoacoustic and photothermal contrast agents with enhanced near-infrared contrast ( approximately 10(2)-fold) for targeting lymphatic vessels in mice using extremely low laser fluence levels of a few mJ cm(-2). Antibody-conjugated golden carbon nanotubes were used to map the lymphatic endothelial receptor, and preliminary in vitro viability tests show golden carbon nanotubes have minimal toxicity. This new nanomaterial could be an effective alternative to existing nanoparticles and fluorescent labels for non-invasive targeted imaging of molecular structures in vivo.

Figure 1. AFM and TEM image analyses

a–h, Results for single-walled carbon nanotubes (a–d) and GNTs (e–h). Topographic AFM images (a,e); surface plots (45° view) of magnified topographic images of a and e (b,f); TEM images (c,g); AFM section analysis results (d,h). Scale bars, 100 nm for AFM images (a,e) and 10 nm for TEM images (c,g).

Figure 2. Absorption spectra of carbon nanotubes and GNTs, and PA and PT analyses of GNTs

a, Normalized optical spectra (left vertical axis) of GNTs in ddH₂O (red line), shortened single-walled carbon nanotubes in ddH₂O (black line) and ddH₂O only (green line) and normalized PA signal amplitudes (right vertical axis) of GNTs in ddH₂O (blue circles) at different laser wavelengths. The concentration of carbon nanotubes was ~35 times higher than that of GNTs. b–j, PT images (b–d), PT thermolens signals (e–g) and PA signals (h–j) from GNTs under a single laser pulse at 850 nm. The laser fluence and time delay were 10 mJ cm⁻² and 5 ns (b), 55 mJ cm⁻² and 5 ns (c) and 0.2 J cm⁻² and 70 ns (d). Amplitude (vertical axis), timescale (horizontal axis) and laser fluence were 25 mV, 1 ms div⁻¹ and 20 mJ cm⁻² (e), 200 mV, 1 ms div⁻¹ and 60 mJ cm⁻² (f), 500 mV, 2 ms div⁻¹ and 0.2 J cm⁻² (g), 50 mV, 2 ms div⁻¹ and 15 mJ cm⁻² (h), 100 mV, 1 ms div⁻¹ and 55 mJ cm⁻² (i), and 500 mV, 2 ms div⁻¹ and 0.2 J cm⁻² (j). k, PA signal amplitudes from GNTs as a function of laser fluence. Inset: detailed linear and nonlinear signal behaviour at low laser fluences. l, PA signals from GNTs as function of laser pulse number at 850 nm and under various laser fluences. The error bars represent the standard deviation in five measurements.

Figure 3. Comparison of PA signals and bubble thresholds among various nanoparticles

a,b, PA signal amplitudes (a) and bubble-formation thresholds (b) of spherical gold nanoparticles (GNPs, 40 nm in diameter), single-walled carbon nanotubes (SWNTs, 1.7 nm× 186 nm), gold nanorods (GNRs, 15 nm×52 nm), gold nanoshells (GNSs, 140-nm silica core with 8-nm gold shell), and GNTs (11 nm × 99 nm) in ddH₂O at a concentration of 1× 10¹¹ particles ml⁻¹ and under a single laser pulse. Laser fluence was ~100 mJ cm⁻² (a). Laser wavelengths were 532 nm for spherical GNPs and 850 nm for the remainder of the nanoparticles. The error bars represent the standard deviation in ten measurements. *P<0.05, compared to GNTs.

Figure 4. Schematics of GNT-assisted PA/PT molecular diagnostics and therapeutics

a, GNT synthesis and its delivery to the target. b, Principle of targeting endothelial LYVE-1 receptors with antibody–GNT complex (left panel), and PA (top right) and PT (bottom right) detection schematics.

Figure 5. In vivo molecular targeting of murine lymphatics with GNTs guided by an integrated PA/PT technique

a, Fragment of mouse mesentery with mapping area (square, 280 mm_ 280 mm). b, Enlarged version of i. c, Laser-induced localized (~10 µm in diameter) lymphatic wall damage. d–k, PT (d–g) and PA (h–k) two-dimensional lymphatic mapping in selected mesenteric area before GNT administration (d,h), at 60 min after administration of antibody–GNTs (e,i) and at 15 min (f,j) and 60min (g,k) after administration of GNTs alone. Dashed white lines in d–k indicate the lymphatic wall and valve. Arrows in b indicate valve leaflets. Total scanning time was 16 s. Laser parameters: wavelength, 850 nm; fluences, 35 mJ cm⁻² (d,h), 10 mJ cm⁻² (e–g, i–k) and 60 mJ cm⁻² (c).