Thermal ablation of tumor cells with antibody-functionalized single-walled carbon nanotubes

Abstract

Single-walled carbon nanotubes (CNTs) emit heat when they absorb energy from near-infrared (NIR) light. Tissue is relatively transparent to NIR, which suggests that targeting CNTs to tumor cells, followed by noninvasive exposure to NIR light, will ablate tumors within the range of NIR. In this study, we demonstrate the specific binding of antibody-coupled CNTs to tumor cells in vitro, followed by their highly specific ablation with NIR light. Biotinylated polar lipids were used to prepare stable, biocompatible, noncytotoxic CNT dispersions that were then attached to one of two different neutralite avidin-derivatized mAbs directed against either human CD22 or CD25. CD22(+)CD25(-) Daudi cells bound only CNTs coupled to the anti-CD22 mAb; CD22(-)CD25(+) activated peripheral blood mononuclear cells bound only to the CNTs coupled to the anti-CD25 mAb. Most importantly, only the specifically targeted cells were killed after exposure to NIR light.

Fig. 1.

Water-soluble CNTs functionalized with biotinylated polar lipids. (a) AFM image of B-CNTs shows CNTs coated by the biotinylated polar lipid, DSPE-PEG-biotin. (b) TEM images of individual B-CNTs show uniform coverage of biotin after immunodetection with gold-labeled anti-biotin. (Inset) Higher magnification of a B-CNT coated with gold-labeled antibiotin. (c) UV-Vis-NIR spectrum of B-CNTs show a number of metallic and semiconducting CNT absorbances consistent with the presence of individual tubes. (d) Raman spectra of B-CNTs show an intense G band (≈1,590 cm⁻¹) indicating the presence of CNTs. In all cases, one representative experiment of at least three independent experiments is shown.

Fig. 2.

Analysis of mAb-NA conjugates. (a) A typical chromatographic separation of RFB4-NA from unconjugated RFB4 and NA using a Sephacryl S-300 HR column. Fractions of the first peak containing the RFB4-NA conjugate were pooled and concentrated. (Inset) Purified RFB4-NA, RFT5-NA, or mAb were electrophoresed under nondenaturing conditions on a 7.5% polyacrylamide gel and immunoblotted with HRP-labeled sheep anti-mouse IgG. Data in a are representative of at least three independent experiments. (b) A total of 5 × 10⁴ Daudi cells were incubated for 24 h with increasing amounts of the RFB4-NA conjugate, and cytotoxicity was detected by [³H]thymidine incorporation. Similar concentrations of unconjugated RFB4 or NA were used as negative controls, whereas 10 μg/ml goat anti-IgM was used as positive control. Data represent mean ± SD of three independent experiments. (c) One million Daudi cells precoated with a saturating concentration of RFB4-NA were incubated with increasing amounts of B-CNT. Saturating concentration of RFB4-NA can target 0.237 pg of B-CNT per Daudi cell. No detectable B-CNT binding was found on uncoated cells or cells precoated with RFT5-NA (control). Data represent mean ± SD of three independent experiments.

Fig. 3.

Optical properties of CNTs following coupling with mAbs (mAb-CNT). (a) UV-Vis-NIR spectrum of RFB4-CNTs show the same metallic and semiconducting CNT types as observed for the B-CNTs, indicating the retention of the optical properties of CNTs after the coupling with RFB4-NA. The sharp feature at 861 nm is caused by a grating and detector change associated with the spectrometer. (b) Raman spectrum of RFB4-CNTs show an intense G band (≈1,590 cm⁻¹) as the B-CNTs, indicating the presence of CNTs in the conjugate. The spectra are representative of three independent experiments.

Fig. 4.

Binding of mAb-CNTs to target cells. One million cells were incubated with saturating concentrations of RFB4-CNTs or RFT5-CNTs and then incubated either with FITC-GAMIg to detect the mAbs or with PE-SA to detect the B-CNTs and analyzed on a FACScan. (a) The specific binding of RFB4-CNTs to Daudi cells using RFT5-CNTs as a negative control (P < 0.001). (b) The specific binding of RFT5-CNT to activated PBMCs (>95% T cells) using RFB4-CNT as a negative control (P < 0.002). Data represent mean ± SD of at least three independent experiments.

Fig. 5.

Ablation of mAb-CNT-coated cells with NIR. One million cells were incubated with saturating concentrations of RFB4-CNTs or RFT5-CNTs. Cells were dispensed into 96-well plates, exposed for 7 min to 808-nm NIR light (5 W/cm²), pulsed with 1 μCi [³H]thymidine, and harvested 12 h later. The incorporated radioactivity was measured by liquid scintillation counting from triplicate samples. The percentage of radioactivity incorporated by each sample was calculated relative to corresponding nonirradiated sample. (a) The specific killing by RFB4-CNTs of Daudi cells using RFT5-CNTs as a negative control (P < 0.0001). (b) The specific killing of RFT5-CNT on activated PBMCs (>95% T cells) using RFB4-CNTs as a negative control (P < 0.0001). Data represent mean ± SD of at least three independent experiments. (c) The stability of the mAb-CNTs in vitro was determined by incubating them in mouse serum at 37°C for 0, 24, 48, and 72 h. At each time point, the mAb-CNTs were washed with PBS, incubated with Daudi cells, and exposed to NIR light as described above. The activity of the RFB4-CNTs at the different time points remained unchanged. *, P < 0.0001; **, P < 0.05 for the values obtained at the corresponding time points with RFT5-CNTs. Data represent mean ± SD of three independent experiments.