MRI of breast tumor initiating cells using the extra domain-B of fibronectin targeting nanoparticles

Abstract

The identification of breast tumor initiating cells (BTICs) is important for the diagnosis and therapy of breast cancers. This study was undertaken to evaluate whether the extra domain-B of fibronectin (EDB-FN) could be used as a new biomarker for BTICs and whether EDB-FN targeting superparamagnetic iron oxide nanoparticles (SPIONs) could be used as a magnetic resonance imaging (MRI) contrast agent for BTIC imaging in vitro and in vivo. BTICs (NDY-1) exhibited high EDB-FN expression, whereas non-BTICs (MCF-7, BT-474, SUM-225, MDA-MB-231) did not exhibit EDB-FN expression. Furthermore, Cy3.3-labeled EDB-FN specific peptides (APTEDB) showed preferential binding to the targeted NDY-1 cells. To construct an EDB-FN targeted imaging probe, APTEDB was covalently attached to a thermally cross-linked SPION (TCL-SPION) to yield APTEDB-TCL-SPION. In the in vitro MRI of cell phantoms, selective binding of APTEDB-TCL-SPION to NDY-1 cells was evident, but little binding was observed in MCF-7 cells. After the intravenous injection of APTEDB-TCL-SPION into the NDY-1 mouse tumor xenograft model, a significant decrease in the signal within the tumor was observed in the T2*-weighted images; however, there was only a marginal change in the signal of non-targeting SPIONs such as APTscramble-TCL-SPION or TCL-SPION. Taken together, we report for the first time that EDB-FN was abundantly expressed in BTICs and may therefore be useful as a new biomarker for identifying BTICs. Our study also suggests that APTEDB-TCL-SPION could be used as an MRI contrast agent for BTIC imaging.

Keywords: Aptides; Breast tumor initiating cells; Extra domain-B of fibronectin; Magnetic resonance imaging.; Superparamagnetic iron oxide nanoparticles.

Figure 1

Analysis of the genes expressed in breast tumor initiating cells (BTICs). (A) RT-PCR analysis of the self-renewal genes and surface marker genes in BTICs (NDY-1) and breast cancer cells (MCF-7). Specific markers of BTICs and self-renewal genes were highly expressed in NDY-1 cells but not in MCF-7 cells. (B) Flow cytometry analysis of the surface markers CD44 and CD24. NDY-1 exhibited a CD44+/CD24- BTIC phenotype. (C) ALDEFLUOR assay for ALDH1 activity and Immunostaining analysis for ALDH1 expression. Approximately 26% of the NDY-1 cells exhibited brightly fluorescent ALDH1 staining, indicating the presence of a putative BTIC marker. Immunostaining showed NDY-1 spheroids strongly expressed the ALDH1 protein. Scale bar: 20 μm

Figure 2

Analysis of the expression of extra domain-B of fibronectin (EDB-FN) and its selective targeting with an EDB-FN aptide (APT_EDB). (A) RT-PCR analysis of the EDB-FN mRNA in diverse breast cancer cell lines. EDB-FN was expressed in NDY-1 cells but not in SUM-225, MCF-7, BT-474 or MDA-MB-231 cells. (B) EDB-FN targeting images of the Cy3.3-labeled APT_EDB (Cy3.3-APT_EDB, red fluorescence) and immunostained EDB-FN (green fluorescence) images in breast cancer cells. Cy3.3-APT_EDB specific signals were observed in EDB-FN overexpressing NDY-1 cells but not in MCF-7 cells. Scale bar: 20 μm

Figure 3

Synthetic procedures and characterization of APT_EDB-TCL-SPION. (A) Schematic illustration of the preparation of APT_EDB-TCL-SPIONs. (B) A TEM image of APT_EDB-TCL-SPION. Scale bar: 20 nm (C) T₂-weighted images and plots of the relaxation rates (R₂ =1/T₂) of APT_EDB-TCL-SPION, APT_scramble-TCL-SPION, and TCL-SPION at various iron concentrations. The transverse relaxivities (r₂) of APT_EDB-TCL-SPION, APT_scramble-TCL-SPION and TCL-SPION were 232.1 ± 1.4 mM^-1 sec^-1, 211.1± 2.1 mM^-1 sec^-1, and 259.9 ± 3.2 mM^-1 sec^-1, respectively.

Figure 4

In vitro MRI analysis of cell phantoms. (A) Representative T₂-weighted images of cell phantoms. Cells were treated with APT_EDB-TCL-SPION and APTscramble-TCL-SPION (11.2 µg Fe/ml) for 12 h at 37°C. Blocking of specific binding of APT_EDB-TCL-SPION was achieved by pretreatment with EDB-FN aptides (APT_EDB, 0.1 mg/ml) for 1 h. Hypointense signals (arrow) were clearly detected in NDY-1 cells treated with APT_EDB-TCL-SPIONs. (B) R₂ values measured from the T₂-weighted images obtained from cell phantoms. The R₂ values significantly increased in the APT_EDB-TCL-SPION-treated NDY-1 cells compared with the APT_scramble-TCL-SPION-treated or untreated NDY-1 cells. Significantly different R₂ values were not observed in APT_EDB-TCL-SPION- and APT_scramble-TCL-SPION- treated and untreated MCF-7 cells. Pre-incubation with free APT_EDB inhibited the increase in R₂ values observed for APT_EDB-TCL-SPION-treated NDY-1 cells. All values are presented as the mean ± standard deviation of at least three independent experiments. Asterisks (*) indicate that the p value was statistically significant (<0.05).

Figure 5

In vivo EDB-FN targeted tumor MRI. (A) T₂*-weighted multi-slice images of the NDY-1 tumor in the mouse obtained prior to injection and at 4 h and 24 h after the injection of APT_EDB-TCL-SPION or APT_scramble-TCL-SPION (20 mg Fe/kg). The dotted line with circles indicates the engrafted tumor region. Multifocal hypointense spots were observed in tumors obtained from mice injected with APT_EDB-TCL-SPION. (B) The signal intensity changes of the tumor areas, liver, spleen and kidney obtained from T₂*-weighted images. An apparent signal intensity decrease within the tumor was detected only after APT_EDB-TCL-SPION injection. All values are presented as the mean ± standard deviation of at least three independent experiments. Asterisks (*) indicate that the p value was statistically significant (<0.05).

Figure 6

Histological analysis of the tumors. (A) Hematoxylin and eosin (H&E) staining, Prussian blue staining and EDB-FN immunostaining. Prussian blue staining showed that a larger number of accumulated SPIONs were detected as blue dots in the tumors obtained from mice injected with APT_EDB-TCL-SPION compared with APT_scramble-TCL-SPION. Immunostaining of EDB-FN in NDY-1 tumor microsections was performed using the BC-1 antibody. EDB-FN proteins (dark brown) were abundantly detected in the NDY-1 tumors injected with APT_EDB-TCL-SPION or APT_scramble-TCL-SPION. (B) Immunostaining of CD31. Immunostaining of the platelet endothelial cell adhesion molecule (CD31) showing that vessels (dark brown) were observed in NDY-1 tumors obtained from mice injected with APT_EDB-TCL-SPION. (C) Co-staining of an anti-EDB-FN antibody and Prussian blue. The accumulation of APT_EDB-TCL-SPION (blue dots) was observed in EDB-FN positive areas of the tumor vasculature and interstitium. Scale bar: 100 μm