Preparation and characterization of FMN
Fluorescent magnetic nanobeads were prepared by an oil-in-water emulsion-evaporation technique (Figure 1). In this approach, poly(styrene-co-maleic anhydride) (PSMA) was selected as the matrix material for nanobeads because PSMA is initially oil-soluble and then can be transferred into a hydrophilic surface by hydrolysis of anhydride groups into carboxyl groups after nanobeads formation. Oil-soluble QDs and MNPs dispersed in PSMA polymer chloroform solution used as the oil phase. Subsequently, an oil/water miniemulsion was formed through stirring and ultrasonication. The gradual removal of chloroform from the aqueous phase by evaporation was compensated by transferring chloroform from the mini droplets to the aqueous phase. Finally, the polymer precipitated, entrapped QDs and MNPs inside polymer nanobeads, and formed polymer/nanoparticle hybrid nanobeads.
The TEM image of the prepared FMN, as shown in Figure 2A, demonstrates that the FMN has a diameter in the range of about 40-95 nm with a broad size distribution. However, using the dynamic light scattering technique, the average hydrodynamic diameter of FMN was 114 nm and a polydispersity index (PDI) of 0.13 (Figure 2B), indicating the relatively monodisperse and stability of the prepared system. Due to the mild reaction process and the polymer matrix isolation between QDs and MNPs, QDs encapsulated in the polymer nanospheres still possess high fluorescent intensity. The digital image of FMN suspension (Figure 2C) was highly fluorescent under UV light illumination. It can be rapidly isolated (in 10 min) from the solution to the tube wall using an external magnet.
From the magnetic hysteresis loop characterization (Figure 2D), the FMN had an excellent superparamagnetic property at room temperature with a tunable magnetic saturation value (41.6 emu/g and 6.04 emu/g) by adjusting the concentration of magnetic nanoparticles added in the oil phase. The high magnetic saturation FMN was used for the construction of iFMNS. To apply the FMN for CTCs isolation and detection, as shown in Figure1, streptavidin was covalently coated on the surface of FMN by a classic carboxyl-amino reaction under the activation of carbodiimide. Then biotinylated anti-EpCAM antibody was attached to the surface by the streptavidin-biotin interaction just before cell isolation.
Tumor cell capturing using iFMNS
The iFMNS (FMN + anti-EpCAM antibody) concentration applied for cell capture was preoptimized using a suspension of SGC-7901 cells. The results are shown in Figure 3A. With the increase of iFMNS concentration, the capture efficiency increased until the concentration reached 0.2 mg/mL, and 96% of SGC-7901cells were captured. Then, the magnetic attraction time was investigated, as shown in Figure 3B, nearly 100% of tumor cells were captured in approximately 8 min. Finally, we used 0.3 mg/mL iFMNS under 10 min of magnetic isolation for further experiments. The magnetic isolation of cells was respectively performed using FMN-SA and iFMNS as capturing probes. As shown in Figure 3C, nearly no cancer cells (3%) were captured by FMN-SA indicated that our iFMNS (with anti-EpCAM antibody) was specifically binding to cancer cells. Moreover, the iFMNS were also applied for several other cancer cell line isolation, as shown in Figure 3D. The EpCAM-negative lymphoblast Raji cells were not isolated by magnetic separation, which also indicated the specificity of the prepared iFMNS. The cell capture efficiencies were more than 70% in various kinds of cancer cells, showing that iFMNS had general applicability for EpCAM-positive cancer cells.
Then, the capability of iFMNS to capture tumor cells to mimic CTC blood samples was investigated. The samples were prepared by spiking green cell trace dye pre-stained SGC-7901 cells into whole blood. The captured cells were enumerated under a fluorescence microscope. The tumor cells and red blood cells were effectively separated from each other using magnet separation in a centrifuge tube, as shown in Figure 4A. Although there were still some red blood cells in the isolated tumor cell suspension mainly due to the gravitational sedimentation, the isolation impurity can be improved primarily by using a magnetic column.
For comparison of the matrix effect, CTC capture efficiencies were also examined in PBS, whole blood, and lysed blood spiked with CTC at concentrations of5-1000 cells/mL. The results are shown in Figure 4B. The capture efficiencies in the three types of samples were comparable and did not have significant differences. These results suggested that complex conditions had negligible effects on the binding between iFMNS and the target cells, and iFMNS could be directly used in whole blood.
Identification of the captured cells
The isolated CTCs were identified based on fluorescent labeling and morphology. With the ultra-bright fluorescent intensity of iFMNS and quick response magnetic separation performance, we investigate the specific recognition and labeling of iFMNS towards tumor cells. As shown in Figure 4C, the green cell trace pre-stained SGC-7901 cells were specifically isolated. The EpCAM motif on the surface of SGC-7901 cells can be directly visualized by the red fluorescent signals from the as-prepared iFMNS. While the different expression amount of EpCAM can also be semi-quantified through fluorescent intensity from iFMNS on the cell surface.
Besides, FITC labeled anti-CD45 (CD45-FITC, green) stained with WBCs and DAPI (blue) stained nuclear. The three-color was used to discriminate tumor cells from WBCs. CTCs are defined as a positive signal for EpCAM (iFMNS, red), DAPI stained nuclear and negative signal for CD45-FITC, while WBCs are defined as a negative signal for iFMNS, positive signal for DAPI, and CD45-FITC. Figure 5 shows that the intense red and blue fluorescence were observed within DAPI and iFMNS stained SGC-7901 cells. Meanwhile, no fluorescence signal was observed for the channel of CD45-FITC. As a control, WBC cells, which are CD45 positive and EpCAM negative, are easily stained with green fluorescence but without red fluorescence. These observations describe that CTCs are specifically recognized in the blood samples from WBCs. The results confirm that the as-constructed system could efficiently isolate and identify CTCs to mimic clinical samples.
Cell viability and function of the captured cells
The viability of isolated tumor cells was investigated to evaluate this isolation method's influence on the cell for further culture and analysis. The viability of cancer cells was measured by trypan blue dyeing. As shown in Figure 6A, the dead cells would be stained by trypan blue dyes, while the live cells would not be stained. The viability rate of isolated tumor cells was calculated to be about 95%, indicating that most of the tumor cells are live after magnetic isolation. Apoptosis analysis of tumor cells before and after magnetic isolation was performed by FITC-annexin V/PI staining and flow cytometry. As shown in Figure 6B, the proportion of apoptotic cells, including early and late, for the control group was 9.75%, while for captured tumor cells was 5.97%. These results indicated that magnetic separation using iFMNS would not induce the tumor cells apoptosis. The slightly lower proportion of apoptosis in captured tumor cells may attribute to the low expression of EpCAM in apoptosis tumor cells, which would not be captured by iFMNS. Then, the captured SGC-7901 tumor cells were re-cultured and propagated in vitro. As shown in Figure 6C, the iFMNS captured tumor cells proliferated without a significant change in behavior and morphology compared with the control samples of SGC-7901 cell lines. Interestingly, cadmium-containing quantum dots are limited in biological labeling mainly due to toxicity for cell growth, but as-prepared iFMNS had nearly no influence on the cell growth and apoptosis be attributed to the thick polymer shell encapsulation around QDs to prevent cadmium ions leakage.
Moreover, we also investigated the possibility of captured tumor cells for nucleic acid molecular analysis. Reverse transcriptase-PCR, which is widely used for precision tumor patient identification and target therapy, was performed to amplify epidermal growth factor receptor (EGFR) mRNA (Figure 6D) with Glyceraldehyde 3-phosphate Dehydrogenase (GAPDH) as a housekeeping gene (Figure 6E). Figure 6D shows that the 214-bp DNA band of the EGFR coding region was found with the HCC827 cells captured by iFMNS (lane 3 in Figure 6D), while no band was found when iFMNS were used to treat tumor cells (lane 2 in Figure 6D). These suggested iFMNS immunological binding had a negligible influence on RNA extraction and PCR reaction, and RT-PCR can analyze the isolated cells without disassociating iFMNS. Overall, the tumor cells captured with iFMNS had nearly no influence on tumor cells' viability and were suitable for subsequent molecular biological analysis, which was crucial for further clinical diagnosis and research.
CTC capture and testing in patient blood samples
To verify our method can be used for clinical samples, the peripheral blood from a lung cancer patient was collected, and the detection of CTC was performed in 2 days. As shown in Figure 7, intense red and blue fluorescence was observed from iFMNS and DAPI with a tumor cell. Meanwhile, no green fluorescence signal was observed in channel CD45-FITC. For captured cells that are CD45 positive and EpCAM negative with DAPI stained nucleus were identified as white blood cells. The results demonstrated that the as-prepared iFMNS could be used to simultaneously isolate and identify CTCs in real clinical samples with high efficiency.