Identifying EGFR-Expressed Cells and Detecting EGFR Multi-Mutations at Single-Cell Level by Microfluidic Chip
- 1.3k Downloads
KeywordsEGFR mutation Single-cell analysis Microfluidic chip Tyrosine kinase inhibitor
Discovering not only the existence of specific EGFR multi-mutations occurred in minority of EGFR-mutated cells which may be covered by the noises from majority of un-mutated cells, but also other valuable single-cell-level information: on which specific cells the mutations occurred, or whether different mutations coexist on the same cells.
Trapping and identifying EGFR-expressed single cells to exclude interferences from EGFR-unexpressed cells.
Epidermal growth factor receptor (EGFR) has been proved to be related with the pathogenesis and progression of multiple carcinoma types, including lung cancer , breast cancer , prostatic cancer  and pancreatic cancer . Previous clinical trials demonstrated that inhibitors of EGFR tyrosine kinase (TK) effectively retarded disease progression of non-small cell lung cancer (NSCLC) patients [5, 6]. Evidences suggest that mutated EGFR proteins are inhibited by small-molecule tyrosine kinase inhibitors (TKIs) which compete with ATP binding to the TK domain of the receptor and block signal transduction . Mutations mediate oncogenic effects by altering downstream signaling and anti-apoptotic mechanisms [1, 7]. For instance, L858R in exon 21 and Del E749-A750 in exon 19 mutations increase the TKIs sensitivity , while T790M in exon 20 is a drug-resistant mutation, abrogating inhibitors binding with EGFR [9, 10]. Since these mutations significantly affect the effectiveness of targeted medicine, EGFR analysis is becoming more and more a routine test before selecting targeted therapy for related cancers, such as NSCLC [11, 12, 13].
Immunohistochemistry of tumor tissue is the most clinically used method to detect EGFR at protein level [14, 15]. Also, directly sequencing cells extracted from tumor tissue has also been clinically accepted to detect EGFR mutation sequences [16, 17]. However, either the protein analysis or the gene sequencing of tumor tissue provides only averaged information of the whole cell population. Since the tumor cells are heterogeneous [18, 19], the mutations occurred on a small amount of cells could be covered by the other normal cells .
To reveal EGFR mutation on individual cells, fluorescence-activated cell sorting (FACS) was previously introduced  to sort single cells from a large cell amount, usually larger than 105 cells . For cell samples fewer than 105 cells, the emerging microfabrication technologies have advanced the examinations of protein expression or gene mutation at single-cell level by preciously controlling single cells and their surrounding environments. At protein level, by employing immunofluorescence identification, microfluidic chips are capable of identifying [23, 24] or enumerating  EGFR-expressed cells. However, the application of protein level analyses is limited by the diverse specificity of different antibodies and the lack of detailed mutation information. At gene level, on-chip single-cell isolation, lysis and gene amplification have been realized using microchambers  or droplets , enabling the sequencing of the disease-related gene fragments [28, 29] or even the whole genome . However, the lack of on-chip identification of EGFR expression and corresponding sorting of EGFR-expressed cells compromises the feasibility of selectively sequencing EGFR-expressed cells which possibly make up a small portion of all cells extracted from tumor tissue.
Clinically, before performing targeted therapy, it is crucial to understand not only if EGFR expression happens but also how many types of disease-related mutation exist and what the mutated sequences exactly are . This urgent demand is yet to be fulfilled with an accurate, simple and cost-effective method, despite the advances which have already been achieved on EGFR mutation determination, with or without the assistance of microfluidic chips. To address this requirement, we developed a simple microfluidic chip to simultaneously finish on-chip cell identification and in situ cell lysis for detecting EGFR multi-mutations at single-cell level. The on-chip cell identification distinguished EGFR-expressed cells from EGFR-unexpressed cells, providing direct and accurate information about the portion of EGFR-expressed cells. Also, by sequencing only EGFR-expressed cells, the interference from EGFR-unexpressed cells was excluded. The in situ cell lysis ensured the accuracy of DNA sequence by avoiding cross-contamination between different cells and possible cell loss while transferring cells between on-chip and off-chip. After optimizing the operation of the microfluidic chip, we evaluated its performance with NSCLC cells. The results demonstrated that the microchip accurately distinguished NSCLC cells from normal cells and determined three important drug-related EGFR mutations that the NSCLC cells possessed.
3 Experimental Section
3.1 Materials and Cells
Dulbecco’s modified eagle’s medium (DMEM), fetal bovine serum (FBS), penicillin–streptomycin and trypsin were purchased from Life Technologies, USA. Phosphate-buffered saline (PBS, pH 7.4) was purchased from Sigma-Aldrich, USA. EGFR monoclonal antibodies conjugated with fluorescein isothiocyanate (anti-EGFR-FITC) and epithelial cell adhesion molecule monoclonal antibodies conjugated with fluorescein isothiocyanate (anti-EpCAM-FITC) were both purchased from Abcam, USA. The nuclear dye, 4′,6-diamidino-2-phenylindole (DAPI), was purchased from Sigma-Aldrich, USA. Silicon wafers were purchased from Xilika, China. Polydimethylsiloxane (PDMS) was obtained from Dow Corning, USA. Multiple displacement amplification (MDA) REPLI-g single-cell kits were purchased from Qiagen, Germany. Cell lysis buffer and polymerase chain reaction (PCR) kits were purchased from Tiangen, China.
The non-small cell lung cancer cell line NCI-H1975 and NCI-H1650 were cultured in 1640 medium with 1% penicillin–streptomycin and 10% fetal bovine serum (FBS). Non-small cell lung cancer cell line A549, breast cancer cell line MCF-7 and human embryonic kidney cell line HEK-293T were cultured in DMEM medium with 1% penicillin–streptomycin and 10% FBS. All cells were incubated at 37 °C under 5% CO2 atmosphere. Before experiments, cells were fixed using a 4% paraformaldehyde solution and then labeled by immunofluorescence. All cell lines were stained by DAPI to indicate cell nuclei. MCF-7 and HEK-293T were mixed and stained by Anti-EpCAM-FITC. A549, NCI-H1975, NCI-H1650, and HEK-293T were mixed and stained by anti-EGFR-FITC. Then cells were rinsed three times to exclude excessive fluorescently labeled antibodies.
3.2 Fluorescently Identifying, In Situ Lysing, Amplifying and Directly Sequencing MCF-7 Cells
Forward primer: 5′-TCTAGCAGCAGCTCATGGTG-3′;
Reverse primer: 5′-GGAGCCCAAGGTTCTGAGT-3′.
3.3 Detecting EGFR Multi-Mutations
Exon 19 forward: 5′-AACGTCTTCCTTCTCTCTCTGTCAT-3′
Exon 19 reverse: 5′-CACACAGCAAAGCAGAAACTCAC-3′
Exon 20 forward: 5′-ACCATGCGAAGCCACACTGACGTGCCTCTCCCTCCCTCCAG-3′
Exon 20 reverse: 5′-GTAATCAGGGAAGGGAGATACGGGGAGGGGAGATAAGGAGCCA-3′
Exon 21 forward: 5′-CCCTCACAGCAGGGTCTT-3′
Exon 21 reverse: 5′-GTCTGACCTAAAGCCACCTC-3′
4 Results and Discussion
From the clinical point of view, an ideal technology for detecting EGFR multi-mutations should achieve the following: (1) accurate enough to precisely provide sequence information about specific kinds of mutations; (2) simple and cost-effective to be accepted as a routine test before cancer targeted therapy. To fulfill these requirements, we developed a microwells array-based microfluidic chip to firstly identify EGFR-expressed cells from EGFR-unexpressed cells, then in situ lysis all EGFR-expressed cells for the following gene sequencing.
4.1 The Microfluidic Chip
By analyzing fluorescent images, all EGFR-expressed cells were identified from EGFR-unexpressed and their positions were marked. To lyse all trapped cells, after opening all bottom outlets and closing top outlet, cell lysis solution was pumped into the channel to fill cell trapping array and all cell lysate collecting chambers (schemed in Fig. 1e). Then the bottom outlets were switched off for 30 min until all cells were fully lysed. The cell lysates were maintained in cell lysate collecting chambers through the through-hole at the bottom of each microwell. All through-holes were 8 µm in side length and 170 µm in depth. The through-hole design ensured (as simulated in Fig. S1) all cell lysates were transferred to collecting chamber, without any cross-contamination among different trapping wells.
Finally, the top outlet and all bottom outlets were opened, and cell lysates were retrieved through the bottom outlets, with the assistance of negative pressure, which was generated by an external syringe pump. The square cell lysate collecting chamber (1.5 mm in side length, 1 mm in depth and 2.25 µL in volume) was specially designed to be much larger than the cell trapping chamber. Each lysate collecting chamber covered 100 cell trapping chambers. By controlling the initial cell density, we realized that each lysate collecting chamber contains cell lysates from a few cells (< 4). As long as the ratio between EGFR-mutated and normal cells was more than 1:3, the mutated sequence could be detected by the Sanger’s sequencing method . It meant that we could sequence all cells (< 4) from a chamber, which contained at least 1 EGFR-expressed cell, to detect if any specific mutation exists in EGFR-expressed cells.
In addition, it was easier to retrieve cell lysate from a larger chamber, avoiding the loss of cell lysate and corresponding inaccurate sequencing results. Multiple displacement amplification (MDA) was introduced for unbiased amplification of the whole genome of cell lysates. Depending on how many mutation types needed to be determined, the amplification product was divided into several parts which were, respectively, amplified again by polymerase chain reaction (PCR) with different primers for specific domains. The final amplification products were directly sequenced to reveal specific gene mutations (schemed in Fig. 1f). Compared with previous one-time PCR amplification  in which only one domain could be examined from lysates retrieved from one cell, the combination of MDA and PCR provided the capability of accurately sequencing different domains from the same cell lysate. On the other hand, for the aim of finding out if specific gene mutations exist in EGFR-expressed cells, not accurately sequencing the whole genome of every cell, our design is a practical alternative to expensive deep sequencing of single cells.
4.2 Single-Cell Identification and DNA Sequencing for Detecting EGFR Multi-Mutation
After verifying the feasibility of in situ identifying and lysing few cells on microfluidic chip for sequencing, we tested detecting EGFR multi-mutations on microfluidic chip. To mimic the real clinical samples in which EGFR-expressed cells account for a small portion and different types of mutations coexist in the same sample [35, 36], we mixed A549 cells (EGFR-expressed, wild type), NCI-H1975 cells (EGFR-expressed, point mutation L858R in exon 21 and T790M in exon 20), NCI-H1650 cells (EGFR-expressed, deletion mutation E746-A750 in exon 19) and HEK-293T cells (EGFR-unexpressed) at a ratio of 1:1:1:15. This ratio reflects a typical situation for tumor tissue in which EGFR-expressed cell account for about 10%–20% [35, 36]. A549, NCI-H1975 and NCI-H1650 cells are all NSCLC cells. L858R, T790M and E746-A750 are known as the most important mutations which are directly related to drug responses .
Accurately discovering specific EGFR mutations, especially uncovering the mutation information from a small amount of mutated cells, which could be covered by the noises from other un-mutated cells, is currently becoming an urgent clinical requirement, since several key mutations have proven playing critical roles influencing drug responses of targeted cancer therapies. This requirement is yet to be satisfied with a simple, accurate and cost-effective method. This study provides a microfluidic-chip-based strategy in which the fluorescent identification of EGFR-expressed cells, in situ cell lysis, MDA and PCR gene amplification are integrated to provide high-quality gene amplification products from which the EGFR multi-mutations information could be acquired using simple and low-cost Sanger’s sequencing. This new strategy has the following prominent features: (1) by excluding cells without EGFR expression and limiting the cell numbers of each sequencing to < 4, or even only one cells, the majority of noises which interfere gene sequencing are excluded; therefore, the multi-mutations of a small portion of cells can be detected by simple and cheap Sanger’s sequencing, not expensive deep sequencing; (2) differs from expensive tissue-level NGS or ARMS method which are capable of detecting only the existence of specific mutations, our method provides other valuable single-cell-level information: on which specific cells the mutations occurred, or whether different mutations coexist on the same cells; (3) trapping and lysing single cells in microwells which are isolated from each other eliminate the cross-contamination and cell loss. Also, the combination of MDA and PCR amplification ensures the high quality of gene amplification products for acquiring accurate sequencing results. After optimizing the operation parameters, we verified the new strategy with cell mimics, which contain three most important EGFR mutations. The results reveal that the new strategy is capable of provide the answers of not only if the EGFR expression exists (by fluorescent identification), but also what the mutated sequences exactly are and on which cells these mutations occur.
Overall, for many clinical practices in which EGFR-expressed cells account for a small portion of the whole cell population, this study provides a new method for accurately detecting disease-related EGFR multi-mutations by employing a simple microfluidic chip and the cost-effective Sanger’s sequencing, as an economically affordable alternative to the expensive NGS or ARMS analysis of the whole cell population.
This work was supported by the National High-Tech R&D Program of China (No. 2015AA020408), National Natural Science Foundation of China (No. 61204118, 81500900 and 21503054), Beijing Municipal Science and Technology Project (No. Z171100002017013), Key Research Program of the Chinese Academy of Sciences, Grant NO. KFZD-SW-210.
- 3.M. Prewett, P. Rockwell, R. Rockwell, N.A. Giorgio, J. Mendelsohn, H.I. Scher, N.I. Goldstein, The biologic effects of C225, a chimeric monoclonal antibody to the EGFR, on human prostate carcinoma. J. Immunother. Emphas. Tumor Immunol. 19(6), 419–427 (1996). https://doi.org/10.1097/00002371-199611000-00006 CrossRefGoogle Scholar
- 6.M.G. Kris, R.B. Natale, R.S. Herbst, T.J. Lynch Jr., D. Prager et al., Efficacy of gefitinib, an inhibitor of the epidermal growth factor receptor tyrosine kinase, in symptomatic patients with non-small cell lung cancer: a randomized trial. JAMA 290(16), 2149–2158 (2003). https://doi.org/10.1001/jama.290.16.2149 CrossRefGoogle Scholar
- 10.C.-H. Yun, K.E. Mengwasser, A.V. Toms, M.S. Woo, H. Greulich, K.-K. Wong, M. Meyerson, M.J. Eck, The T790M mutation in EGFR kinase causes drug resistance by increasing the affinity for ATP. Proc. Natl. Acad. Sci. 105(6), 2070–2075 (2008). https://doi.org/10.1073/pnas.0709662105 CrossRefGoogle Scholar
- 12.R. Rosell, E. Carcereny, R. Gervais, A. Vergnenegre, B. Massuti et al., Erlotinib versus standard chemotherapy as first-line treatment for european patients with advanced EGFR mutation-positive non-small-cell lung cancer (eurtac): a multicentre, open-label, randomised phase 3 trial. Lancet Oncol. 13(3), 239–246 (2012). https://doi.org/10.1016/S1470-2045(11)70393-X CrossRefGoogle Scholar
- 13.L.V. Sequist, J.C.H. Yang, N. Yamamoto, K. O’Byrne, V. Hirsh et al., Phase III study of afatinib or cisplatin plus pemetrexed in patients with metastatic lung adenocarcinoma with EGFR mutations. J. Clin. Oncol. 31(27), 3327–3334 (2013). https://doi.org/10.1200/JCO.2012.44.2806 CrossRefGoogle Scholar
- 14.J.R. Grandis, M.F. Melhem, E.L. Barnes, D.J. Tweardy, Quantitative immunohistochemical analysis of transforming growth factor-α and epidermal growth factor receptor in patients with squamous cell carcinoma of the head and neck. Cancer 78(6), 1284–1292 (1996). https://doi.org/10.1002/(SICI)1097-0142(19960915)78:6<1284:AID-CNCR17>3.0.CO;2-X CrossRefGoogle Scholar
- 15.D. Atkins, K.-A. Reiffen, C.L. Tegtmeier, H. Winther, M.S. Bonato, S. Störkel, Immunohistochemical detection of EGFR in paraffin-embedded tumor tissues variation in staining intensity due to choice of fixative and storage time of tissue sections. J. Histochem. Cytochem. 52(7), 893–901 (2004). https://doi.org/10.1369/jhc.3A6195.2004 CrossRefGoogle Scholar
- 17.X. Tang, H. Shigematsu, B.N. Bekele, J.A. Roth, J.D. Minna, W.K. Hong, A.F. Gazdar, I.I. Wistuba, EGFR tyrosine kinase domain mutations are detected in histologically normal respiratory epithelium in lung cancer patients. Cancer Res. 65(17), 7568–7572 (2005). https://doi.org/10.1158/0008-5472.CAN-05-1705 CrossRefGoogle Scholar
- 18.S.K. Singh, I.D. Clarke, M. Terasaki, V.E. Bonn, C. Hawkins, J. Squire, P.B. Dirks, Identification of a cancer stem cell in human brain tumors. Cancer Res. 63(18), 5821–5828 (2003). http://cancerres.aacrjournals.org/content/63/18/5821
- 20.X. Fan, F.B. Furnari, W.K. Cavenee, J.S. Castresana, Non-isotopic silver-stained SSCP is more sensitive than automated direct sequencing for the detection of PTEN mutations in a mixture of DNA extracted from normal and tumor cells. Int. J. Oncol. 18(5), 1023–1026 (2001). https://doi.org/10.3892/ijo.18.5.1023 Google Scholar
- 21.M. Geens, H. Van de Velde, G. De Block, E. Goossens, A. Van Steirteghem, H. Tournaye, The efficiency of magnetic-activated cell sorting and fluorescence-activated cell sorting in the decontamination of testicular cell suspensions in cancer patients. Hum. Reprod. 22(3), 733–742 (2007). https://doi.org/10.1093/humrep/del418 CrossRefGoogle Scholar
- 32.T.K. Yung, K.C. Chan, T.S. Mok, J. Tong, K.F. To, Y.M. Lo, Single-molecule detection of epidermal growth factor receptor mutations in plasma by microfluidics digital pcr in non-small cell lung cancer patients. Clin. Cancer Res. 15(6), 2076–2084 (2009). https://doi.org/10.1158/1078-0432.CCR-08-2622 CrossRefGoogle Scholar
- 34.H.A. Hammond, L. Jin, Y. Zhong, C.T. Caskey, R. Chakraborty, Evaluation of 13 short tandem repeat loci for use in personal identification applications. Am. J. Hum. Genet. 55(1), 175–189 (1994). https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1918216/pdf/ajhg00040-0180.pdf
Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.