Abstract
Cell-in-cell structures (CICs), characterized by the presence of one or more viable cells inside another one, were recently found important player in development, immune homeostasis and tumorigenesis etc. Incompatible with ever-increasing interests on this unique phenomenon, reliable methods available for high throughput quantification and systemic investigation are lacking. Here, we report a flow cytometry-based method for rapid analysis and sorting of heterotypic CICs formed between lymphocytes and tumor cells. In this method, cells were labeled with fluorescent dyes for fluorescence-activated cell sorting (FACS) by flow cytometry, conditions for reducing cell doublets were optimized such that high purity (>95%) of CICs could be achieved. By taking advantage of this method, we analyzed CICs formation between different cell pairs and found that factors from both internalized effector cells and engulfing target cells affect heterotypic CICs formation. Thus, flow cytometry-based FACS analysis would serve as a high throughput method to promote systemic researches on CICs.
Similar content being viewed by others
Introduction
For decades, pathologists have seen a type of special structures, characterized by morphologically normal cells enclosed by other cells, in a variety of human tumor samples. Some terms, such as “cell cannibalism” and “cytophagocytosis” and the like, were used to describe these unique structures, which were recently given a unified name “cell-in-cell” structures (CICs)1,2. Based on the cells involving in structure formation, CICs could roughly be classified into two categories: 1) homotypic CICs, in which structures are formed between cells of same type like epithelial cells; 2) heterotypic CICs, where cells of different types such as epithelia and lymphocytes participate in structure formation2.
Researches on CICs became a subject of interests in recent years, largely due to the finding that formation of CICs would lead to the death of majority of the internalized cells3,4,5,6,7,8. Therefore, one process responsible for homotypic CICs formation, entosis, was recommended as a death mechanism by the Nomenclature Committee on Cell Death9. Recent progresses have shown that CICs formation played important roles in immune homeostasis5 and tumorigenesis10,11,12 and is likely an evolutionarily conserved phenomenon2. Mechanistically, CICs formation may reflect the competitive nature of confronted cells12, representing a novel mechanism of cell competition13.
Boosted interests on CICs call for reliable methods for further investigation. We have previously reported methods for the study of entosis, where homotypic CICs were quantified manually by microscopic observation14. While microscopic counting was accepted for quantification of various CICs, homotypic11,15 or heterotypic5,16, this method turned out to be relatively subjective and labor-intensive and time-consuming for multiple samples. Flow cytometry provides an ideal technique to quantify cells carrying specified fluorophores in a high throughput manner17,18,19. In light of this, we attempt to develop a flow cytometry-based method for CICs analysis. In this study, we demonstrated that heterotypic CICs, formed between tumor cells and lymphocytes, could be identified and sorted out by fluorescence-activated cell sorting (FACS) method under the condition that cell doublets were minimized before flow cytometry analysis. Furthermore, analysis of CICs formed between different cell pairs indentified an active role of host cells in heterotypic CICs formation, which may revise current view that internalizing cells alone drive CICs formation.
Methods
Cell culture and treatment
Cell lines PLC/PRF/5, MCF7, SK-BR-3, RD were purchased from American Type Culture Collection (ATCC, Manassas, VA) and cultured as described20. Molt-4, Raji and BxPC3 were kindly gifted by Prof. Ya-jun Guo (The Second Military Medical School, China) and cultured as described20. NK92MI cells were gifted from Bin Gao (Institute of microbiology Chinese Academy of Sciences) and were grown in α-Modified Eagle Medium (α-MEM) plus 12.5% fetal bovine serum (FBS) and 12.5% horse serum (Gibco BRL, Carlsbad, CA). Cytokine induced killer (CIK) cells were gifted from Wei-dong Han (Chinese PLA General Hospital) and cultured as described4.
Co-culture experiments
Target tumor cell suspension was stained with 2.5 μM CellTracker Green CMFDA dye (Invitrogen, Carlsbad, CA) for 30 min at 37°C in the absence of serum. Monolayer of the tumor cells were incubated in DMEM with 10% FBS at a density of 3.5 × 105 cells/well in 6 well cell culture cluster (Corning, Union City, CA) for 12 h at 37°C in order to adhere. Immune cells were stained with 10 μl CD45-PE (Beckman Coulter, Brea, CA) for 20 min at room temperature prior to co-incubations with adherent tumor cells by a density of 3.5 × 105 cells/well at indicated time to allow CICs formation. Flow cytometry was gated by unstained/stained effector cells and tumor cells alone and unstained effector/tumor cells co-cultures as well.
CICs quantification of Giemsa-stained samples
Unbound cells were removed at the indicated times, adherent cells were washed twice in PBS and fixed in 4% glutaraldehyde for 20 min at room temperature followed by staining with Giemsa (Jiancheng, Nanjing, China). Heterotypic CICs were quantified by a light microscopy (Olympus Optical Co., Tokyo, Japan). The rate of heterotypic CICs was determined as described20 by dividing the number of host cells containing immune cells by total number of tumor cells in a specified field: CICs rate (%) = (number of tumor cells with immune cells/number of total tumor cells counted) × 100. About 800 ~ 1000 tumor cells from random representative fields were counted for each single preparation. Immune cells wrapped at least half-way around by target cells were considered to be internalized. Internalization of multiple immune cells into one tumor cell was counted as one CICs.
FACS analysis of heterotypic CICs by flow cytometry
Since flow cytometry requires cells of non-adherent, cell suspensions were prepared as followed. After co-cultured experiment, remove the medium, wash cells with Dulbecco's Phosphate Buffered Saline (D-PBS) twice and incubated with 0.25% trypsin/EDTA solution (300 μl/well) at 37°C until cells were dissociated from the bottom of the culture dish. Then added same amount of Hank's Balanced Salt Solution (HBSS) with 20% FBS to quench trypsin. Cells were pelleted at 150 × g for 5 min and re-suspended in 400 μl HBSS with 20% FBS. Then the cell suspensions were filtered through flow injection pipe (BD, Cat: 352035) before acquisition on a flow cytometer. Cell-in-cell structures were counted and enriched on a BD FACScan flow cytometer (FACSAria II) (BD Biosciences, San Jose, CA) equipped with a 15 mW, 488 nm, air-cooled argonion laser and three photomultipliers with band pass filters of 530 nm (FL1), 550 nm (FL2) and 670 nm (FL3). CellTracker Green CMFDA was excited by a blue laser and detected by a 530/30 filter. CD45-PE was excited by a blue laser and detected by a 560/40 filter. FSC and SSC voltages of 121 and 324, respectively and a threshold of 2,000 on FSC were applied to gate on the tumor cells population. All observations were made in log mode. Data were acquired by using BD CellQuest PRO software (BD Biosciences, San Jose, CA) and analyzed by using FlowJo flow cytometry analysis software (Tree Star, Ashland, OR).
To establish optimal instrument settings to gate and reduce background noise, effector cells and tumor cells alone were used as single negative control, effector and tumor cells co-cultures were used as double negative control, CD45-PE stained effector cells or CellTracker Green CMFDA stained tumor cells were used as single positive control. CICs frequency was calculated by the equation: CICs (%) = (double-positive cells in quadrant Q2/single-positive tumor cells in quadrant Q3) × 100.
Confocal microscopy
Prepare 300 μl cell suspensions to make cytospin on glass slides in a Cytocentrifuge 7620 (Wescor, Logan, UT) at 400 rpm on high acceleration for 5 min. Cytospins were fixed in 4% glutaraldehyde for 20 min at room temperature. Fixed samples were washed in PBS, applied one drop of antifade reagent with DAPI (Invitrogen, Carlsbad, CA) and mounted on the slides with cover slips followed by sealing with nail oil. Imaging was taken by using an FV1000 laser scanning confocal microscope (Olympus Optical Co., Tokyo, Japan).
Time-lapse microscopy
Time-lapse microscopy was performed as follows. Tumor cells were grown as monolayer on 30 mm glass bottom dishes to adhere for 4 h prior to adding effector cells. Images were obtained every 2 min for the indicated time courses by FV1000 laser scanning confocal microscope (Olympus) and Olympus Image Browser software (Olympus), with a 40× objective and an environmental chamber to maintain temperature (37°C) and CO2 concentration (5%).
Statistics
All experiments were carried out in triplicate and data are indicated as mean ± SD and more than 400 cells were counted in each experiment. GraphPad Prism (GraphPad Software, La Jolla, CA) was used to plot data generated and carry out statistical analysis. P values were calculated by Two-tailed Student's t-test, with statistical significance assumed at P < 0.05.
Results
Two-color assay to analyze heterotypic CICs by FACS
Heterotypic CICs formed between tumor cells and lymphocytes were usually quantified by microscopic counting of Giemsa-stained co-cultures16,20. Although this method is simple and convenient, the results were sometimes affected by homotypic CICs formed between tumor cells, as it's hard to clearly tell the identity of the internalized cells, especially when they are dying. Moreover, this method was less efficient in analyzing multiple samples and only limited number of cells could be given. One way to solve these problems is to stain the two types of cells in fluorescently different colors and analyze co-cultured cells by flow cytometry. In light of this, we designed a flow cytometry-based method as depicted in Figure 1. Briefly, the lymphocytic effector cells (suspended NK92MI) and epithelial target tumor cells (attached PLC/PRF/5) were stained with CellTracker CMFDA in green and anti-CD45-PE in red respectively before co-culturing for several hours to allow CICs formation. NK92MI cells that don't bind to or penetrate into PLC/PRF/5 cells were washed out prior to preparing cell suspensions for flow cytometry by fully trypsinization. Theoretically, double-positive cells (in Q2) could be read as the heterotypic CICs formed between lymphocytes and tumor cells, while those of single positive are cells not forming CICs. In this way, CICs could be analyzed in a high throughput manner.
Suppressing cell doublet formation by EDTA, DNase and fetal bovine serum
Our pioneer experiments indicated that flow cytometry couldn't differentiate CICs from double-positive doublets, which sometimes constituted the majority of double-positive population (data not shown). In order to control doublet formation, we optimized the concentration of EDTA and DNase, two factors potentially affecting cell-cell adhension, in HBSS used for cell suspension. As shown in Figure 2A and 2B, although less cell doublet formation was observed in HBSS with higher concentration of EDTA or DNase, considerable amount of cell doublets (about 7% for both) could still form in very high EDTA or DNase condition, which might potentially be toxic to cells and therefore are not ideal for doublet inhibition. Interestingly, we found that doublet formation was efficiently suppressed to quite low level (about 1.5%) when cells were suspended in HBSS with 20% fetal bovine serum (FBS) (Figure 2C), probably due to its ability to block the binding sites for cell-cell adhesion. Accordingly, double-positive cells, including doublets and CICs, decreased drastically from ~12.5% to ~4% with increased FBS concentrations (Figure 2D), indicating the presence of large amounts of cell doublets in double-positive cell population in low-serum medium. A combination of 20% FBS with EDTA and/or DNase didn't give significantly better inhibition of cell doublets (data not shown), therefore, 20% FBS is used for following analysis.
CICs analysis by FACS correlates well with that by microscopic counting
As described above, high concentration of FBS might be a good candidate to inhibit doublet formation, which will help read out CICs from double-positive cell population. To test this, we performed FACS analysis of cells prepared in 20% FBS. Double-positive cells were sorted out and observed by microscopy. As shown in Figure 3A, CICs were highly enriched after FACS, the typical morphology is that each green cell (CMFDA) is surrounding a black hole filled with red dots (CD45-PE) and blue nucleus (red arrows), an average purity of >85% could be generally achieved with some experiments gave >95% purities (Figure 3E). Moreover, after 4-hours co-culture of different effector:target ratio, the CICs frequencies quantified by FACS (Figure 3B) were tightly correlated (r2 = 0.9918) with those from manual microscopic counting (Figure 3C), suggesting CICs constitutes majority part of the double-positive cell population when cells were prepared in 20% FBS HBSS, which is a suitable condition for CICs analysis by flow cytometry.
Intriguingly, we found that the percentage of CICs with typical morphology among CICs sorted by FACS decreased with longer co-culture (Figure 3E), although the overall frequency increased with time (Figure 3D). We reasoned this for death and fragmentation of internalized effector cells, generating atypical CICs characterized by the presence of pieces of red dots in host target cells (yellow arrows in Figure 3A). In agreement with this, we found, by microscopic time-lapse, that some internalized cells underwent rapid death and fragmentation, morphologically resembling those atypical CICs sorted from flow cytometry, within 6 hours (Figure 4).
FACS is also applied to analyze other heterotypic CICs
To examine whether the FACS method is also applicable to analyze heterotypic CICs formed between a broad ranges of lymphocytes and tumor cells, we introduce into this assay additional cell lines, including four additional tumor cells (MCF-7, SK-BR-3, BxPC3 and RD) and three more types of immune cells including T cells (Molt-4), B cells (Raji) and cytokine induced killer cells (CIK). Heterotypic CICs formed between them were systemically analyzed by flow cytometry along with manual microscopic counting side by side. In all cases, the CICs quantifications from FACS displayed no significant differences from those by microscopic counting (Figure 5), suggesting FACS by flow cytometry is a reliable method to detect heterotypic CICs formed between lymphocytes and tumor cells. Interestingly, we found that some tumor cells, such as PLC/PRF/5, displayed higher abilities over others to form CICs with immune cells, which suggests that host/tumor cells, in addition to effector/immune cells, also play important role in CICs formation as discussed below.
Discussion
Although being extensively reported in human tumors for a long time, CICs were short of mechanistic studies until recent years. Several forms of research models had been established, including entosis3, emperitosis4, heterotypic cell cannibalism (HeCC)6, homotypic cell cannibalism (HoCC)8, suicidal emperipolesis (SE)5 and phagoptosis21 and the like. While homotypic CICs formed by mechanisms like entosis and HoCC mainly involve in tumor development and progression8,10,11,12, heterotypic CICs formed by emperitosis, HeCC and SE etc. are implicated in broader biological processes, such as development of immune system22, immune homeostasis5, tumorigenesis4,16 and neurodegeneration23 and the forth. Our previous work showed that heterotypic CICs were actually prevalent in inflammatory tissues20. Therefore, it's important for the field to develop an efficient method for accurate and high throughput analysis of heterotypic CICs. In this study, we established a method based on flow cytometry technique for rapid analysis of heterotypic CICs (Figure 1) and conditions for discriminating CICs from cell doublets were optimized to obtain high purity of CICs readout (averagely > 85%) (Figure 2 and 3). In addition, this method was shown to be also useful to analyze CICs of various cell pairs (Figure 5), thus providing a powerful tool for further mechanistic study of heterotypic CICs. It should be noted that traditional microscope-based analysis of CICs could not distinguish partial and complete cell-cell internalizations. As we know, due to cell-cell interaction, many partial cell-cell internalizations do not eventually become real CICs. The method described here utilizes Trypsin to dissociate the cells, under which condition the partial cell-cell internalization will be disrupted. Therefore, this method is helpful to analyzes genuine CICs.
For the assay to be successful, in addition to reducing doublets while maintaining cells viability, particular attention should be paid to the way to label cells fluorescently. The fluorophores must be highly sensitive and not leaky. We use CellTracker Green CMFDA and CD45-PE to stain tumor cells and immune cells, respectively. The advantage of CellTracker Green CMFDA and CD45-PE are their shared stimulation by laser of 488 nm that is more standard across all ranges of instrumentations, including less expensive benchtop models that use solid state lasers. Other dyes such as CellTracker Orange CMTMR and CellTracker Red CMTPX belong to this series of fluorescent probes, but need special excitation wavelength and didn't work well for flow cytometry assay in our hands (data not shown). We choose CD45-PE on the account of its specificity for immune cells and no tumor cell staining. Furthermore, the probe and antibody didn't show detectable influence on cell viability and proliferation which is necessary for further research. Accordingly, the combination of CellTracker Green CMFDA and CD45-PE is recommended for analyzing heterotypic CICs between tumor cells and CD45-positive lymphocytes.
Distinct from phagocytosis of dying cells, CICs formation involves the engulfment of viable cells. Moreover, in most cases, the engulfment was believed to be an active process driven by the internalizing cells rather than the cells being penetrated3,16. And polarized actomyosin contraction compartmentalized by junctional p190A RhoGAP within internalizing cells, together with epithelial cadherins-mediated adherens junctions, promotes CICs formation in entosis11. Nevertheless, the outer host cells were also found playing roles during CICs formation in some cases. For example, metastatic melanoma cells displayed higher ability than primary tumor cells to engulf both dead and live lymphocytes6 and Nupr1 down-regulation conferred pancreatic tumor cells ability to engulf their neighbors upon activating TGF-β signaling8. In this work, we systemically analyzed the CICs formed between a group of tumor cell-lymphocyte pairs by using the methods described above. Interestingly, we found that lymphocytes differed in their abilities to penetrate different tumor cells to form heterotypic CICs. For example, NK92MI cells could form CICs with PLC/PRF/5 in a frequency of 12% or so in 4 hours, the frequency is less than 1% for RD cells, which is consistent across all the tumor cell-lymphocyte pairs examined (Figure 5). This phenotype could be explained by a role of host target cells (tumor cells) in CICs formation, at least in this work. Taking together our data and those published, we proposed that both internalizing effector cells and host target cells participated in the formation of CICs, which is an integral part of current view that CICs formation is actively driven by the internalizing cells. Future researches on the topic will provide more mechanistic interpretation for this phenotype.
References
Overholtzer, M. & Brugge, J. S. The cell biology of cell-in-cell structures. Nat Rev Mol Cell Biol 9, 796–809, 10.1038/nrm2504 (2008).
Wang, X. Cell-in-cell phenomenon: a new paradigm in life sciences. Curr Mol Med (2015) [In press].
Overholtzer, M. et al. A nonapoptotic cell death process, entosis, that occurs by cell-in-cell invasion. Cell 131, 966–979, 10.1016/j.cell.2007.10.040 (2007).
Wang, S. et al. Rapid reuptake of granzyme B leads to emperitosis: an apoptotic cell-in-cell death of immune killer cells inside tumor cells. Cell Death Dis 4, e856, 10.1038/cddis.2013.352 (2013).
Benseler, V. et al. Hepatocyte entry leads to degradation of autoreactive CD8 T cells. Proc Natl Acad Sci U S A 108, 16735–16740, 10.1073/pnas.1112251108 (2011).
Lugini, L. et al. Cannibalism of live lymphocytes by human metastatic but not primary melanoma cells. Cancer Res 66, 3629–3638, 10.1158/0008-5472.CAN-05-3204 (2006).
Wang, Y. & Wang, X. Entosis and related forms of cell death within Cells. Curr Mol Med (2015) [In press].
Cano, C. E. et al. Homotypic cell cannibalism, a cell-death process regulated by the nuclear protein 1, opposes to metastasis in pancreatic cancer. EMBO Mol Med 4, 964–979, 10.1002/emmm.201201255 (2012).
Galluzzi, L. et al. Molecular definitions of cell death subroutines: recommendations of the Nomenclature Committee on Cell Death 2012. Cell Death Differ 19, 107–120, 10.1038/cdd.2011.96 (2012).
Krajcovic, M. et al. A non-genetic route to aneuploidy in human cancers. Nat Cell Biol 13, 324–330, 10.1038/ncb2174 (2011).
Sun, Q., Cibas, E. S., Huang, H., Hodgson, L. & Overholtzer, M. Induction of entosis by epithelial cadherin expression. Cell Res 24, 1288–1298, 10.1038/cr.2014.137 (2014).
Sun, Q. et al. Competition between human cells by entosis. Cell Res 24, 1299–1310, 10.1038/cr.2014.138 (2014).
Huang, H., Chen, Z. & Sun, Q. Mammalian cell competitions, cell-in-cell phenomenon and their biomedical implications. Curr Mol Med (2015) [In press].
Sun, Q. & Overholtzer, M. Methods for the study of entosis. Methods Mol Biol 1004, 59–66, 10.1007/978-1-62703-383-1_5 (2013).
Wan, Q. et al. Regulation of myosin activation during cell-cell contact formation by Par3-Lgl antagonism: entosis without matrix detachment. Mol Biol Cell 23, 2076–2091, 10.1091/mbc.E11-11-0940 (2012).
Wang, S. et al. Internalization of NK cells into tumor cells requires ezrin and leads to programmed cell-in-cell death. Cell Res 19, 1350–1362, 10.1038/cr.2009.114 (2009).
Galbraith, D. W., Anderson, M. T. & Herzenberg, L. A. Flow cytometric analysis and FACS sorting of cells based on GFP accumulation. Methods Cell Biol 58, 315–341 (1999).
Tung, J. W. et al. Modern flow cytometry: a practical approach. Clin Lab Med 27, 453–468, v, 10.1016/j.cll.2007.05.001 (2007).
Carter, A. D., Bonyadi, R. & Gifford, M. L. The use of fluorescence-activated cell sorting in studying plant development and environmental responses. Int J Dev Biol 57, 545–552, 10.1387/ijdb.130195mg (2013).
Chen, Y.-H. et al. Prevalence of heterotypic tumor/immune cell-in-cell structure in vitro and in vivo leading to formation of aneuploidy. PLoS One 8, e59418, 10.1371/journal.pone.0059418 (2013).
Brown, G. C. & Neher, J. J. Eaten alive! Cell death by primary phagocytosis: ‘phagoptosis'. Trends Biochem Sci 37, 325–332, 10.1016/j.tibs.2012.05.002 (2012).
Guyden, J. C. & Pezzano, M. Thymic nurse cells: a microenvironment for thymocyte development and selection. Int Rev Cytol 223, 1–37 (2003).
Brown, G. C. & Neher, J. J. Inflammatory neurodegeneration and mechanisms of microglial killing of neurons. Mol Neurobiol 41, 242–247, 10.1007/s12035-010-8105-9 (2010).
Acknowledgements
The authors thank Dr. Shengqi Wang for kindly suggestion and support, Dr. Qianqian Dai for technical advice and discussion. This work was supported by the National Basic Research Program of China (2015CB553704, Q.S.), the National Natural Science Foundation of China (81472588, Q.S.; 81273197, 81471578, X.N.W.; 81172337, W.L.; 81302689, Z.L.C.), the Pioneer Research Foundation, Beijing Institute of Biotechnology (Q.S.), Pioneer Project for Young Scientist of Military Biomedicine (Q.S.), the Open Project from National Key Laboratory of Medical Immunology (Q.S.), Beijing Municipal Natural Science Foundation (7142121, K.T.Y.) and Beijing Nova Program (Z141107001814013, Q.H.L.).
Author information
Authors and Affiliations
Contributions
Q.S. conceived the project. M.H. conducted most experiments with help from H.H. and M.W. A.C., X.N., K.Y. and Q.L. provided assistance in imaging and data analysis. W.L., L.M., Z.C. and X.W. provided scientific guidance. M.H. and Q.S. primarily analyzed the data. Q.S. wrote the manuscript with help from M.H. All authors reviewed the manuscript.
Ethics declarations
Competing interests
The authors declare no competing financial interests.
Rights and permissions
This work is licensed under a Creative Commons Attribution 4.0 International License. The images or other third party material in this article are included in the article's Creative Commons license, unless indicated otherwise in the credit line; if the material is not included under the Creative Commons license, users will need to obtain permission from the license holder in order to reproduce the material. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/
About this article
Cite this article
He, M., Huang, H., Wang, M. et al. Fluorescence-Activated Cell Sorting Analysis of Heterotypic Cell-in-Cell Structures. Sci Rep 5, 9588 (2015). https://doi.org/10.1038/srep09588
Received:
Accepted:
Published:
DOI: https://doi.org/10.1038/srep09588
This article is cited by
-
Tumor malignancy by genetic transfer between cells forming cell-in-cell structures
Cell Death & Disease (2023)
-
Cell-in-cell phenomenon: leukocyte engulfment by non-tumorigenic cells and cancer cell lines
BMC Molecular and Cell Biology (2021)
-
Hydrophoresis — A Microfluidic Principle for Directed Particle Migration in Flow
BioChip Journal (2020)
-
Impaired formation of homotypic cell-in-cell structures in human tumor cells lacking alpha-catenin expression
Scientific Reports (2015)
-
Implication of cell-in-cell structures in the transmission of HIV to epithelial cells
Cell Research (2015)
Comments
By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.