3.1. Separation and Analysis of the Different Lung Cancer Cell Lines on the Single-Cell Microarray Chip
To investigate the possibility of analyzing single nucleotide-mutated cancer cells at the single cell level, we analyzed the ratio or the exact number of gene-mutated cancer cells using the combined technologies of single-cell microarray chips and PNA-DNA probes. As a model of anticancer drug-resistant cancer cells, a lung cancer cell line, NCI-H1975, which harbors the T790M mutation, the resistance factor against anticancer drugs, was screened from multiple non-mutated lung cancer cell lines, A549. The single-cell microarray chip used in this study had 62,410 microchambers, which had an upper diameter of 31–40 µm, a lower diameter of 11–20 µm, a depth of 28 µm, and a pitch of 100 µm (Figure 1
a,b). In our previous study [7
], the NCI-H1650 lung cancer cell line was successfully separated into single cells using microchambers with an upper diameter of 32 µm and a lower diameter of 12 µm of one cluster in the chip. As the NCI-H1975 and A549 cell lines used in this study were approximately the same size as the NCI-H1650 cell line, microchambers with a 32-µm upper diameter and a 12-µm lower diameter were used for separating them into single cells. Dispersing the cancer cell suspension into the single-cell microarray chip resulted in NCI-H1975 and A549 occupying approximately 50% and 60% of microchambers on the chip, respectively (Figure 2
). The two types of lung cancer cells were successfully separated into single cells and aligned on a single-cell microarray chip in the same way. These results indicate that, even if the two types of cancer cells (single nucleotide-mutated and non-mutated cells) are mixed, it is possible to separate them and determine the exact ratio of target mutation-harboring cancer cells on the single-cell microarray chip.
This study used DAPI for staining the cellular nucleus, PE-labeled anti-CK antibody for staining the CK, and an FITC-conjugated PNA-DNA probe for staining the T790M-mutated EGFR mRNA. PNA has high target binding ability due to PNA’s lack of electronic charge in the backbone and has the high specificity [49
]. These PNA probes are expected to be more suitable than DNA probes for gene mutation analysis of single-nucleotide mutations. In a previous study [48
], we developed the PNA-DNA probes for as an easy and rapid detection for EGFR gene mutations, such as exon19del E746-A750, T790M, and L858R. The developed PNA-DNA probes showed increases in fluorescence intensity against dose-dependent, in vitro target EGFR-mutated sequences, which indicates that FITC-PNA probes are separated from Q-DNA probes, and specifically bind to the target sequence of mutated EGFR mRNA (Figure 1
b). To analyze the T790M-mutated cancer cell on the single-cell microarray chip, staining dyes were introduced into the cancer cells using the surfactant saponin after the cell suspension was dispersed onto the cell chip. The fluorescence of the stained cancer cells was observed under a confocal-laser scanning microscope. The DAPI and CK antibodies displayed a similar fluorescence intensity in every cancer cell. As antibody or nuclear staining could be successfully performed on the single-cell microarray chip, it was demonstrated that multiple staining was also possible on the microarray chip. The mean FITC fluorescence intensity in single cells was calculated using imaging software HCImage (Hamamatsu Corporation, Bridgewater, NJ, United States). The fluorescence intensities of NCI-H1975 (T790M-mutated cell) and A549 (non-mutated cell) were 116.6 and 53.5, respectively (Figure 2
). We concluded that the mutation-harboring cancer cells could be stained and discriminated on a single-cell microarray chip because the NCI-H1975 cells had a fluorescence intensity twice that of the negative control (A549) cells. As the maximum fluorescence intensity of A549 was approximately 80–100, the cells with a fluorescence intensity of greater than 100 in this study were classified as positive (NCI-H1975) cells, and the cells with a fluorescence intensity of 100 or lower were classified as A549 cells. The differences of fluorescent intensity of NCI-H1975 and A549 cell lines were enough to discriminate each other in single-cell microarray chip under the fluorescence microscope combined with PNA-DNA probe technology. Therefore, these results strongly indicated the imaging analysis of single nucleotide-mutated single-cancer cells can be achieved. To our knowledge, there have been no reports confirming that lung cancer cells containing the mutated and non-mutated cells have been aligned and analyzed in single cells using a microarray. The single-cell microarray chip technology in this research has made it possible to easily and accurately separate and analyze the mutation-harboring cancer cells at the single cell level. Various cancer cells can be separated into single cells, regardless of their type, suggesting that single-cell microarray chip technology could be used for the analysis and diagnosis of a wide range of cancers.
3.2. Detection of T790M-Mutated Cancer Cells Using the Combined Technologies of the Single-Cell Microarray Chip and the PNA-DNA Probe
For gene mutation analysis using the most versatile system, the NGS system, the system generally requires more than 20% of cells within the sample to have the target mutation [44
]. In this study, to examine the sensitivity of our detection system, 5%, 10%, or 20% NCI-H1975 (T790M-mutated cancer cell line) cells were spiked into A549 cells and dispersed onto a single-cell microarray chip. The NCI-H1975 cell were identified after the cells were dispersed in single-cell microarray chip. The cells were stained by DAPI, CK, and PNA-DNA probes on the chip. All single-cancer cells (mutated and non-mutated) were filled into 50–60% of the microchambers on the single-cell microarray chip and stained with DAPI and CK, without dislodging the cells from the microchambers by the washing process (Figure 3
a). In addition, spiked ratio of cell samples did not appear to change before and after dispersing on the single-cell microarray chip because two types of lung cancer cells were separated into single cells on the chip in the same way after the washing process (Figure 2
). This suggested that multiple stainings at the single-cell level on the single-cell microarray chip had been achieved.
The target T790M-mutated single-cancer cells could be stained by FITC-conjugated PNA-DNA probes depending on the spiked ratio of mutated cancer cells (Figure 3
a). We calculated the fluorescence intensity of the T790M-mutated cells that were observed on the single-cell microarray chip and counted the number of cells having a fluorescence intensity of not less than 100, which is the maximum intensity of the negative control cells (A549). There were 40 ± 7.5 (4.4 ± 0.8%), 68 ± 60 (10.1 ± 8.9%), and 158 ± 42 (21.8 ± 5.8%) positive cells in the 5%, 10%, and 20% NCI-H1975 samples, respectively (Figure 3
b,c). The theoretical cell numbers of NCI-H1975 were 45.7, 68.1, and 144.8 cells in the 5%, 10%, and 20% NCI-H1975 samples, respectively. Therefore, the 5%, 10% and 20% mutant cell samples were exactly proportional to the theoretical ratio. The results show that this system can detect the target mutated single-cancer cells containing 5% or higher ratio. There were no positive cells in the 0% NCI-H1975 (100% A549) samples. Therefore, the positive cells detected in this study had a high probability of being NCI-H1975 cells. Figure 3
c also indicates that each spiked ratio of T790M-mutated cells showed a strong linearity, and the coefficient of determination was 0.998. Some data appear to vary because of the sensitivity, specificity, and cell transduction efficiency of the PNA-DNA probe. It can sometimes cause difficulties in accurate measurement, especially in a low ratio of target single cells. However, improved PNA-DNA probes will enable more sensitive and selective measurement of single nucleotide-mutated cells on single-cell microarray chips. Furthermore, for higher single-cell occupancy and sensitive detection, it is also possible to accurately detect lower ratio of mutated cells by improving the microchamber design of the single-cell microarray chip.
The NGS system, which is currently used in the accurate whole-genome sequencing of cancer cells and identifying anticancer drug-resistant cells, requires expensive equipment and a high level of expertise. In addition, accurate analysis is difficult unless over 20% of cancer cells in the sample contain the target mutation [43
]. Therefore, it is difficult to preemptively diagnose EGFR-mutated cells with NGS using only a few gene-mutated cells collected from the small amount of lung-cancerous tissues. The NGS system generally needs the DNA sample at least 50–1000 ng [43
], which means almost 1.5 × 104
–1.5 × 106
cells are required. On the other hand, the single-cell microarray chips system required total 1 × 106
–5 × 106
cells. The number of cells (sample volume) required for analysis is almost the same level. However, we can further reduce the number of cells by improving the design of microarray chip with higher integrated microchambers. The real-time PCR-based analyzing systems also have high sensitivity, but these conventional methods also require expensive equipment, time-consuming detection (3–5 h for typical PCR systems), and expert technical knowhow. In contrast, the single-cell microarray chip and PNA-DNA probe technology method was able to detect 5–20% of the mutated cells, and the process was completed within 1 h, including cell preparation. This method has been found to be faster and more sensitive than the NGS system because it is easy to use and can be measured using basic fluorescence microscopes. Although the sensitivity of the combined single-cell microarray chip and PNA-DNA probe system is similar to that of several real-time PCR-based methods, it appears that the former system is superior to the latter in terms of detection time and user-friendliness.
Therefore, this combined technology of single-cell microarray chips and PNA-DNA probes system could be a useful tool for analyzing cancer tissue containing a small number of single nucleotide-mutated single-cancer cells collected by biopsy before and after the anticancer drug treatment. In addition, by changing the sequence of PNA-DNA probes to various target mutations, it is possible to detect multiple gene-mutated cancer cells at the single-cell level and effectively select an anticancer agent. A further advantage of the single-cell microarray chip system is the ability to recover target single cells. Previously, we reported the recovery of single-cells by using a micromanipulator and a single-cell microarray chip [7
]. Although we did not recover the T790M-mutated single-cancer cells from the microchambers in the present work, for further analysis it would also be easy to recover the mutated single-cancer cells by using a micromanipulator system in the future.
Recently, minimally invasive diagnostic methods before surgery are strongly required. The liquid biopsies, targeting circulating tumor cell (CTC) or circulating tumor DNA (ctDNA) in the bloodstream are promising diagnostic methods for the prognosis of metastatic cancer. The combined technology of single-cell microarray chips and PNA-DNA probes could also be applicable for the liquid biopsies that require the analysis of a small number of CTCs or ctDNAs from whole blood samples. Although the single-cell microarray chip can detect only 5% target single-cells at this stage, by improving the design of microarray chip with higher integrated microchambers, it will be easy to detect the rare cancer cells like CTCs from larger numbers of cells. Therefore, it is strongly suggested that this method of analysis be used as a new diagnostic tool for cancer by detecting cancer tissue or liquid biopsy samples at single-cell level.