Background

Due to the high transfer ability of cancer cell, cancer has been regarded as a world-wide high mortality disease. It is reported that the increase amount of circulating tumor cells (CTCs) emerged in the peripheral blood will result in the cancer metastasis and relapse [1,2,3]. Quantitative analysis of CTCs can provide some valuable clinical information that is particularly critical for cancer diagnosis and treatment. However, the number of CTCs in the whole bloodstream is usually very low (a few to hundreds per milliliter), quantification of CTCs to assess cancer metastasis faces a huge challenge [4, 5]. In the past few years, various approaches have already been presented to enrich/count CTCs, including flow cytometry [6], immunemagnetic beads [7], microfluidic devices [8,9,10] and so on. Though those developed methods have been greatly accepted, their sensitivity is still a major flaw. Therefore, developing a highly sensitive CTCs assay method is urgent for predicting cancer metastasis and relapse.

Along with the rapid development of micro-/nano-fabrication technique, the three-dimensional (3D) bionic interface-based analysis method has become a hot research topic in the area of nanotechnology and life science. 3D bionic interface, usually displayed in the form of micro-/nano-structure (e.g., micropillars [11], nanofibers [12], nanotubes [13, 14] and nanopillars [15]), are endowed with good biocompatibility and large specific surface area [16,17,18,19]. Simultaneously, combined with the nanoscale characters implanted in cellular surface elements (e.g., microvilli and filopodia) and extracellular matrix (ECM) scaffolds, 3D bionic interface provides a comfortable microenvironment where cell capture and rare cell detection could be achieved. Furthermore, functional nanomaterials have been introduced into 3D bionic interface, significantly diversifying the detection methods and enhancing the detection sensitivity. Wang’s group has ever reported a novel silicon nanopillar coated with anti-EpCAM-based biosensing platform for CTCs capture and subsequent sensitive assay [20]. Our group also developed a graphene-modified 3D microchip-based supersandwich cytosensor for quantitative immunoassay of CTCs [21].

Nanostructure-based devices have been identified as being one of the easiest and most effective techniques for CTC capture applications. Electrospinning is an easy and universal nanofabrication technique, by which a variety of soluble and fusible polymers could be transferred to form the desired nanofibers with steerable diameters from a few nanometers to several micrometers [20, 22,23,24]. The prepared nanofibers are coated onto the 3D bionic interface to simulate an excellent porous microenvironment, which is especially beneficial for cellular filopodia climbing, helping cell adhesion and growth.

Herein, we demonstrated an electrospun nanofibers-deposited nickel (Ni) micropillars-based cytosensor for electrochemical detection of CTCs. Breast cancer cell line with rich EpCAM expression (MCF7) were selected as model CTCs. The ultra-long poly (lactic-co-glycolic acid) (PLGA) nanofibers were firstly-crosswise stacked onto the surface of Ni micropillars by electrospinning to construct a 3D bionic interface for capturing EpCAM-expressing CTCs, following immuno-recognition with quantum dots functionalized anti-EpCAM antibody (QDs-Ab) and forming immunocomplexes on the micro-/nano-chip. The signal current response was achieved by electrochemical assay of the released cadmium ion (Cd2+) after acid-dissolving QDs from immunocomplexes. Using this 3D substrate, we dependably gathered cancer cells from synthetic CTC blood samples. The integration of crossed PLGA nanofibers and conductive Ni micropillars not only provide an excellent microenvironment for CTCs capture, preventing CTCs from flowing away in the process of rinse and increasing the capture efficiency of target cell, but also greatly amplify the current signal, improving detection sensitivity. Coupling with 3D micro-/nano-structure, the proposed biosensing platform exhibited great potential for on-site monitoring cancer progress. We expect that this platform could be applied in isolating rare populations of cells that cannot be easily realized using existing technologies, as well as in early diagnosis and longitudinal monitoring of cancer in the clinic.

Methods

Chemicals

Indium tin oxide (ITO) glasses with a resistance of 10 Ω were purchased from South of China Xiangcheng Technology. AZ9260 photoresists and the developer AZ-300MIF were purchased from AZ Electronic Materials Corp. (Philadelphia, PA). Bovine serum albumin (BSA), fluorescein diacetate (FDA) and streptavidin (SA) were purchased from Sigma (St. Louis, MO). A quantum dots (Qdot) 585 CdSe@ZnS antibody labeling kit was purchased from Life Technologies. DMEM medium for cell culture was obtained from GIBCO. Biotinylated goat IgG polyclonal anti-EpCAM antibody was obtained from R&D Systems (Minneapolis, MN). Pan-Cytokeratin antibody (C11) Alexa Fluor® 488 and CD45 Antibody (2D-1) PE were purchased from Santa Cruz Biotechnology. Cancer patient serum was provided by Zhongnan Hospital of Wuhan University (Wuhan, China). All other chemicals used in this study were analytical-grade. All solutions were prepared with ultrapure water obtained from a Millipore water-purification system (Millipore, USA).

Cell culture

The MCF7 cells were cultured in DMEM, which was supplemented with 10% fetal bovine serum (FBS) and 100 μg/mL penicillin–streptomycin in an incubator (5% CO2, 37 °C). Jurkat cells were cultured in RPMI Medium 1640 supplemented with 10% fetal bovine serum. After the concentration of cells reached 1 × 105 cells mL−1, the cells were collected by centrifugation at 1000 rpm for 3 min respectively.

Fabrication of the 3D Ni micropillars

Scheme 1a–f shows the fabrication procedure of the Ni micropillars on the ITO glass. AZ9260 photoresist was firstly spin-coated onto the ITO glass slide, which was then washed with ultrasonic rinsing in ethanol and distilled water. After exposure with a Chromium mask, which was prepared by Laser Lithography System (Heidelberg, μPG501), beneath UV light, the photoresist film was processed in AZ developer (1:3 v/v AZ-300 MIF/H2O) for 1 min to obtain a desired pattern. Subsequently, the ITO glass slide was immersed in a Ni electroplating solution with the bulk of Ni block as an anode. After electroplating with the current density of 0.05 A/cm2 for 5 min at 50 °C, 3D Ni micropillars with 10 μm height were obtained on the ITO glass slide. Finally, the remaining photoresist was lifted off in acetone.

Scheme 1
scheme 1

Schematic representation of the integrated fabrication of 3D cytosensor and electrochemical detection of CTCs. a–f Fabrication procedure of the Ni micropillars on the ITO glass. g Fabrication of electrospun PLGA nanofibers on Ni micropillars. h CTCs were captured and i QDs-anti-EpCAM antibody conjugates were recognized. j The amplification of captured CTC and k SEM image of CTC in microenvironment. m, n Principles of electrochemical assay

Full size image

Fabrication of electrospinning PLGA nanofibers

Electrospinning PLGA nanofibers were fabricated as follows: PLGA powder (2 g) was liquefied in a mixed solvent of DMF-THF (15 mL, v/v = 3:1) with overnight stirring up to the point when a transparent and homogeneous solution was acquired. Recently assembled PLGA solution filled a stainless steel needle (inner diameter = 0.6 mm) that was connected to a high-voltage DC power supply (Wuhan, China). The electrospinning nanofibers were obtained by using an electrospinning apparatus at a feeding rate of 0.6 mL/h, an electric voltage of 12 kV, and a tip to collector distance of 15 cm. Nanofibers were directly electrospun onto 3D Ni micropillars. The formed 3D micro-/nano-structures were dried in a vacuum oven over 24 h to remove the residual organic solvent and moisture. Finally, Au nanoparticle was plated to the resulted substrate by the ion sputter coater to obtain modified electrode.

Preparation of QDs-anti-EpCAM antibody conjugates

A Qdot 585 antibody labeling kit was utilized to conjugate the anti-EpCAM antibody to QDs. Prior to conjugation, anti-EpCAM was decontaminated via a gel filtration column (Superose 12, Pharmacia-LKB) to eliminate surfactants and other proteins, including BSA. The concentration of anti-EpCAM was about 0.5 mg/mL. Conjugation was performed according to the manufacturer’s procedure. The obtained QDs-anti-EpCAM conjugate was stored at 4 °C for further use. The conjugate concentration was established (1.5 × 106 M) by quantifying the absorbance density at 585 nm with an Ultrospec 2100 Pro UV/visible spectrophotometer.

Quantitative detection MCF7 cells with 3D cytosensor

The MCF7 cell suspension (100 μL) was introduced onto the 3D Ni/PLGA micro-/nano-chip and kept in an incubator (5% CO2, 37 °C) for 1 h. Then, the device was incubated with QDs-anti-EpCAM (100 μL) at 37 °C for 1 h. After rinsing with PBS, a portion of HCl (10 μL, 1 M) was dropped on the chip to dissolve the captured QDs. The detection solution with 10 μg/mL Hg in acetate buffer (0.2 M, pH 4.6) was then added for quantitative measurement with a CHI 660E electrochemical workstation. As a control, substrates only with micropillars or PLGA nanofibers were also examined in parallel.

Results and discussion

Efficient capture of CTCs with the 3D Ni/PLGA micro-/nano-chip

The 3D Ni/PLGA micro-/nano-chip was prepared as illustrated in Scheme 1. As can be seen from Fig. 1a, b, the micropillars were fabricated with the diameter of 40 μm and the height of 10 μm. The diameter of PLGA nanofibers was about 500 nm (Fig. 1c). To test the cell-capture performance of the 3D Ni/PLGA micro-/nano-chip, we fabricated substrates only with Ni micropillars, only with PLGA nanofibers and both with Ni/PLGA, respectively, to compare their cell affinity. A cell suspension (105 cells mL−1) was introduced onto the chips, and then incubated at 5% CO2 and 37 °C for 1 h. As shown in Fig. 1e–g, the Ni/PLGA micro-/nano-chip captured much more cells, while the Ni micropillars surface and PLGA nanofibers surface captured few ones. More important, MCF7 with fully outspread pseudopodia connected to the 3D Ni/PLGA surface could be clearly observed (Fig. 1d). It suggests that the synergistic interplay between Ni and nanofibers could possibly be held accountable for the improved cell-capture yields. This effect likely due to the 3D Ni/PLGA micro-/nano-chip enhanced local topographic interactions among the micropillars, nanofibers and cells.

Fig. 1
figure 1

The 3D Ni/PLGA micro-/nano-chip local topographic interactions among the micropillars, nanofibers and CTCs. The SEM images of a, b Ni micropillars, c PLGA nanofibers and d MCF7 cells were captured on 3D Ni/PLGA micro-/nano-chip. FDA fluorescence micrographs of target cells were captured on e Ni micropillars substrate, f PLGA nanofibers substrate and g 3D Ni/PLGA patterned substrate

Full size image

To differentiate captured CTCs and white blood cells (WBCs), as shown in Fig. 2a–h, three-color immunostaining was performed using Alexa Fluor® 488-labeled anti-CK (Cytokeratin, a protein marker for epithelial cells) and PE-labeled anti-CD45 (a marker for WBCs) as well as Hoechst nuclear staining. Following image capture, the integrated information was used to distinguish CTCs (Hoechst+/CK+/CD45−, cell size > 10 μm) (Fig. 2a) from WBCs (Hoechst+/CK−/CD45+, cell size < 10 μm) (Fig. 2b). On this basis, the mixed suspension consists of CTCs and WBCs were added onto the 3D Ni/PLGA micro-/nano-chip at the same concentration of 105 cells mL−1 and incubated at 37 °C, 5% CO2 for 1 h. After 1× PBS rinse and immunostaining with anti-CK and anti-CD45, it could be observed from Fig. 2c that only CTCs were captured on the 3D substrate. It is mostly due to the size effect that the diameters of the pseudopodia and the PLGA nanofibers are well suited to acquire enough contact and effective adhesive force, causing increased cell/substrate affinity.

Fig. 2
figure 2

The three-color immunocytochemistry technique with Alexa Fluor® 488-labeled anti-CK, PE-labeled anti-CD45 and Hoechst nuclear staining was used to distinguish a CTCs from b non-specifically trapped WBCs on the 3D micro-/nano-chip. c Isolation of CTCs from mixed cell suspension on 3D bionic interface

Full size image

Optimization of experimental conditions

The electrochemical activity of the cytosensor is impacted by several parameters, including the portion of QD-anti-EpCAM conjugate attached to the MCF7 cells on the surface of the 3D Ni/PLGA micro-/nano-chip. It directly affects the electrochemical response of the cytosensor since the signal mainly depends on the amount of QD-anti-EpCAM conjugates recognized by CTCs. The surplus of Qdot conjugates results in increasingly nonspecific adsorption. To acquire a maximum reaction and a minimum nonspecific adsorption, the optimal amount of QD-anti-EpCAM was determined. We diluted the original conjugate solution to various concentrations and investigated the influence on the signal-to-noise (S/N) ratio of the cytosensor with 105 cell mL−1. PBS acted as a control. As shown in Fig. 3a, the S/N ratio was discovered to be the highest for dispensing 100 nM QD-anti-EpCAM. Nevertheless, the reduction of the S/N ratio at a concentration > 100 nM is because of the rise of the background signal causing surplus portions of Qdot conjugates while at a lower concentration, the reduction of the S/N ratio is attributed to the lowering of the signal due to low Qdot-anti-EpCAM availability. Consequently, 100 nM Qdot-anti-EpCAM was typically utilized as the optimal concentration during the whole study.

Fig. 3
figure 3

Optimization of parameters of the 3D cytosensor. a Qdot-anti-EpCAM conjugates concentration on the signal-to-noise (S/N) ratio of cytosensor. b The relationship among the electrospinning time, the PLGA nanofiber density and MCF7 capture density. c Immunoreaction time on the signal responses of cytosensor. They were performed by applying a 105 cell mL−1 to the substrate

Full size image

In addition, in order to investigate how PLGA nanofiber density affects the cell capture density, we performed cell-capture experiments with MCF7 cells on a series of the alternately 3D Ni/PLGA micro-/nano-substrate (PLGA nanofiber with electrospun time of 2, 3, 4, 5, 6 min). As shown in Fig. 3b, the cell capture density increased with increasing electrospun time from 2 to 4 min. When the electrospun time was beyond 4 min, the cell capture densities were reduced, which is likely since too packed nanofiber limited the cell capture. These results reveal that the local topographic interactions are correlated with the characteristics of the PLGA nanofiber.

The incubation time between CTCs and QDs-Ab conjugates has also been explored by electrochemical response, as shown in Fig. 3c. The current reaction was elevated with the rising immunoreaction time, and it was inclined to be stable after about 60 min. Thus, 60 min was selected as the optimized reaction time.

Electrochemical detection of CTCs

Under the optimal experiment condition, the proposed supersandwich cytosensor was challenged to test different concentrations of MCF7 cells. As depicted in Fig. 4, the electrochemical response increased with the amount of MCF7 cells increasing, showing a good linear range of 101–105 cells mL−1 with a correlation coefficient of 0.9937 (n = 3). The detection limit at a signal-to-noise ratio of 3σ—where σ is the standard deviation of the signal in a blank solution—declined to 8 cells mL−1, which greatly benefited from the large specific surface area of electrospun PLGA nanofiber, excellent conductivity of Ni micropillar and signal amplification of QDs. The comparison of the study about CTCs detection methods was summarized in Table 1.

Fig. 4
figure 4

Differential pulse voltammetry responses of the supersandwich cytosensor incubated with (a) Jurkat cells at 105 cells mL−1 and (b–f) different concentrations of MCF7 cells: 101, 102, 103, 104, 105 cells mL−1. Insert: Calibration curve of MCF7

Full size image
Table 1 Comparison of the sensitivity of different CTCs detection methods
Full size table

A series of human plasma samples were further used to test the accuracy and viability of the proposed approach. These samples were established by spiking various amounts of MCF7 to human plasma. The results were summarized in Table 2, which showed the recoveries are in the range of 93.5–105%, indicating that the constructed device provides a novel technique for fast, selective, and sensitive detection of MCF7 in actual specimens. In addition, CTC-capture study was performed on peripheral blood samples from gastric and lung cancer patients. The peripheral blood samples were donated by two types of cancer patients with different stages of the disease and preserved in blood collection tubes, and the results were summarized in Fig. 5. Serum sampling was conducted based on ethics principles established by the biosafety committee at Zhongnan Hospital of Wuhan University and each of the donors signed the consent forms.

Table 2 Comparison of cytosensor values with known amounts of MCF7 spiked in human plasma
Full size table
Fig. 5
figure 5

Detection of CTC amounts in the 1 mL blood samples collected from cancer patients. Green columns: gastric cancer patients. Red columns: lung cancer patients. The amounts were calculated from the calibration curve

Full size image

Conclusions

In summary, we have tactfully fabricated a cancer cell capture and analysis 3D micro-/nano-chip, which achieves effective capture and sensitive quantitation of CTCs mediated by electrochemical assay. The considerable capture performance benefit from a 3D bionic interface provided by Ni micropilliars and electrospinning PLGA nanofiber net where cancer cells can easily adhere. Meanwhile, a supersensitive analysis method is worked out by electrochemical detection Cd2+ released from QDs using an aliquot of HCl. More importantly, using these 3D micro-/nano-chips, we accurately gathered cancer cells from synthetic CTC blood samples and from whole blood samples obtained from gastric and lung cancer patients.