Development and in vitro study of a bi-specific magnetic resonance imaging molecular probe for hepatocellular carcinoma

Abstract

Background: Hepatocellular carcinoma (HCC) ranks second in terms of cancer mortality worldwide. Molecular magnetic resonance imaging (MRI) targeting HCC biomarkers such as alpha-fetoprotein (AFP) or glypican-3 (GPC3) offers new strategies to enhance specificity and help early diagnosis of HCC. However, the existing iron oxide nanoparticle-based MR molecular probes singly target AFP or GPC3, which may hinder their efficiency to detect heterogeneous micro malignant HCC tumors < 1 cm (MHCC). We hypothesized that the strategy of double antibody-conjugated iron oxide nanoparticles which simultaneously target AFP and GPC3 antigens may potentially be used to overcome the tumor heterogeneity and enhance the detection rate for MRI-based MHCC diagnosis.

Aim: To synthesize an AFP/GPC3 double antibody-labeled iron oxide MRI molecular probe and to assess its impact on MRI specificity and sensitivity at the cellular level.

Methods: A double antigen-targeted MRI probe for MHCC anti-AFP-USPIO-anti-GPC3 (UAG) was developed by simultaneously conjugating AFP andGPC3 antibodies to a 5 nm ultra-small superparamagnetic iron oxide nanoparticle (USPIO). At the same time, the singly labeled probes of anti-AFP-USPIO (UA) and anti-GPC3-USPIO (UG) and non-targeted USPIO (U) were also prepared for comparison. The physical characterization including morphology (transmission electron microscopy), hydrodynamic size, and zeta potential (dynamic light scattering) was conducted for each of the probes. The antigen targeting and MRI ability for these four kinds of USPIO probes were studied in the GPC3-expressing murine hepatoma cell line Hepa1-6/GPC3. First, AFP and GPC3 antigen expression in Hepa1-6/GPC3 cells was confirmed by flow cytometry and immunocytochemistry. Then, the cellular uptake of USPIO probes was investigated by Prussian blue staining assay and in vitro MRI (T2-weighted and T2-map) with a 3.0 Tesla clinical MR scanner.

Results: Our data showed that the double antibody-conjugated probe UAG had the best specificity in targeting Hepa1-6/GPC3 cells expressing AFP and GPC3 antigens compared with single antibody-conjugated and unconjugated USPIO probes. The iron Prussian blue staining and quantitative T2-map MRI analysis showed that, compared with UA, UG, and U, the uptake of double antigen-targeted UAG probe demonstrated a 23.3% (vs UA), 15.4% (vs UG), and 57.3% (vs U) increased Prussian stained cell percentage and a 14.93% (vs UA), 9.38% (vs UG), and 15.3% (vs U) reduction of T2 relaxation time, respectively. Such bi-specific probe might have the potential to overcome tumor heterogeneity. Meanwhile, the coupling of two antibodies did not influence the magnetic performance of USPIO, and the relatively small hydrodynamic size (59.60 ± 1.87 nm) of double antibody-conjugated USPIO probe makes it a viable candidate for use in MHCC MRI in vivo, as they are slowly phagocytosed by macrophages.

Conclusion: The bi-specific probe presents enhanced targeting efficiency and MRI sensitivity to HCC cells than singly- or non-targeted USPIO, paving the way for in vivo translation to further evaluate its clinical potential.

Keywords: Alpha-fetoprotein; Glypican-3; Hepatocellular carcinoma; Magnetic resonance imaging; Molecular imaging; Ultra-small superparamagnetic iron nanoparticles.

Figure 1

Flow of cell-based experiments. First, AFP and GPC3 antigen expression on Hepa1-6/GPC3 cells was confirmed by flow cytometry and immunocytochemistry (step 1). Next, the cellular uptake of USPIO probes was investigated by performing Prussian blue-staining assays for iron (step 2). Finally, in vitro MRI was performed, including T2-weighted imaging (T2WI) and T2 Map imaging (step 3). HRP: Horseradish peroxidase; PE: Phycoerythrin; Ab: Antibody; USPIO: Ultra-small superparamagnetic iron oxide; MRI: Magnetic resonance imaging; T2WI: T2-weighted imaging.

Figure 2

Physical characterization of ultra-small superparamagnetic iron oxide probes by transmission electron microscopy and magnetic resonance imaging. A-D: Transmission electron microscopy (TEM) images of ultra-small superparamagnetic iron oxide (USPIO) probes of U, UA, UG, and UAG, respectively. E: The core size distribution of USPIO with a mean diameter of 4.88 nm and a standard deviation of 0.16 nm (n = 147), as determined from the TEM images. F: T2-weighted magnetic resonance images of a series of water solutions containing different concentrations of USPIO as indicated by iron concentration (left) and linear regression fitting of the transversal relaxation rate (1/T2) data vs different iron concentrations for extracting the transverse relaxivity r₂ (right). UAG: Anti-AFP–USPIO–anti-GPC3; UA: Anti-AFP–USPIO; UG: Anti-GPC3–USPIO; U: Unlabeled (non-targeted) USPIO.

Figure 3

Hydrodynamic size distribution of antibody-conjugated ultra-small superparamagnetic iron oxides. A: Schematic illustration of the conjugation between antibodies (anti-AFP and anti-GPC3) and USPIO-NHS ester to form single or double antibody-conjugated USPIO probes. B: Hydrodynamic size distribution of USPIO (U), anti-AFP–USPIO (UA), anti-GPC3–USPIO (UG), and anti-AFP–USPIO–anti-GPC3 (UAG). USPIO: Ultra-small superparamagnetic iron oxide; AFP: Alpha-fetoprotein; GPC3: Glypican-3; NHS ester: Succinimidyl ester.

Figure 4

Detection of alpha-fetoprotein and glypican-3 antigen expression in Hepa1-6/GPC3 cells by flow cytometry. Flow cytometry data showed significantly higher alpha-fetoprotein expression in the cytoplasm (A) than in the membrane (B), compared with blank cell and IgG isotype controls. C: The positive shift of fluorescence distribution compared with isotype control illustrated higher membrane expression of the glypican-3 antigen. AFP: Alpha-fetoprotein; GPC3: Glypican-3.

Figure 5

Cellular immunocytochemistry results. From left to right: Horseradish peroxidase-based immunological staining with IgG isotype control, anti-alpha-fetoprotein antibody, and anti-glypican-3 antibody.

Figure 6

Prussian blue staining of Hepa1-6/GPC3 cells treated with four kinds of ultra-small superparamagnetic iron oxide probes. Prussian blue-staining images of blank Hepa1-6/GPC3 cells (A) and Hepa1-6/GPC3 cells treated with 50 µg Fe/mL of (B) USPIO (U), (C) anti-AFP–USPIO (UA), (D) anti-GPC3–USPIO (UG), or (E) anti-AFP–USPIO–anti-GPC3 (UAG). (F) Quantitation of the percentages of blue stained cells. The total counted cell number for the U, UA, UG, and UAG groups was 84, 133, 199, and 119, respectively. USPIO: Ultra-small superparamagnetic iron oxide; AFP: Alpha-fetoprotein; GPC3: Glypican-3; UAG: Anti-AFP–USPIO–anti-GPC3; UA: Anti-AFP–USPIO; UG: Anti-GPC3–USPIO; U: Unlabeled (non-targeted) USPIO.

Figure 7

In vitro magnetic resonance imaging results demonstrating the binding efficiency and imaging properties of different ultra-small superparamagnetic iron oxide probes. A: T2WI of different samples contained in a 96-well plate. From left to right: H₂O, 1% agar, blank Hepa1-6/GPC3 cells, and hepa1-6/GPC3 cells treated with 100 µg Fe/mL USPIO (U), anti-AFP–USPIO (UA), anti-GPC3–USPIO (anti-GPC3–USPIO), or anti-AFP–USPIO–anti-GPC3 (UAG). B: Pseudocolor T2 map of cell samples treated with U, UA, UG, and UAG, respectively, compared with blank cells. T2 values are illustrated with a color bar. C: Signal intensities of cells after different probe treatments under different TE and exponential fits for the T2 values. Inset: The fitted T2 relaxation time for cells treated with different USPIO probes (blank control, U, UA, UG, or UAG). USPIO: Ultra-small superparamagnetic iron oxide; AFP: Alpha-fetoprotein; GPC3: Glypican-3; UAG: Anti-AFP–USPIO–anti-GPC3; UA: Anti-AFP–USPIO; UG: Anti-GPC3–USPIO; U: Unlabeled (non-targeted) USPIO; T2WI: T2-weighted imaging.