FISH-Flow, a protocol for the concurrent detection of mRNA and protein in single cells using fluorescence in situ hybridization and flow cytometry

Abstract

We describe a flow-cytometry-based protocol for intracellular mRNA measurements in nonadherent mammalian cells using fluorescence in situ hybridization (FISH) probes. The method, which we call FISH-Flow, allows for high-throughput multiparametric measurements of gene expression, a task that was not feasible with earlier, microscopy-based approaches. The FISH-Flow protocol involves cell fixation, permeabilization and hybridization with a set of fluorescently labeled oligonucleotide probes. In this protocol, surface and intracellular protein markers can also be stained with fluorescently labeled antibodies for simultaneous protein and mRNA measurement. Moreover, a semiautomated, single-tube version of the protocol can be performed with a commercially available cell-wash device that reduces cell loss, operator time and interoperator variability. It takes ∼30 h to perform this protocol. An example of FISH-Flow measurements of cytokine mRNA induction by ex vivo stimulation of primed T cells with specific antigens is described.

Figure 1

Comparison of mRNA expression analysis by microscopy and FISH-Flow. (a) Schematic representation of sm-FISH probes tiled along the length of an mRNA target. Binding of multiple fluorescently labeled probes to each mRNA molecule results in fluorescent signals that are bright enough to enable detection by flow cytometry and by fluorescence microscopy. (b) Representative microscopy images of HeLa cells after sm-FISH using c-fos probes. Cells were cultured on 0.17-mm-thick glass coverslips coated with gelatin and subjected to serum treatment as described in the Results section, hybridized with c-fos probes and imaged according to the previously described adherent cell protocol. The presented images are merged optical slices. Scale bars, 5 µm. Images were acquired using a Zeiss Axiovert 200 M microscope equipped with a 100× objective. (c) Scatter plots of HeLa cells after hybridization with probes against c-fos mRNA, GFP mRNA (a negative control) and GAPDH mRNA (a positive control). Cells were either left untreated or treated with serum, detached, resuspended and processed according to the PROCEDURE. Frequencies of cells positive for c-fos, GFP (negative control) and GAPDH mRNA (positive control) are reported above each gate. Gates were set separately, based on no-serum controls in each case. Cy5 fluorescence was recorded using a 640-nm laser and a 670/30 BP filter. A total of 50,000 events per sample were analyzed.

Figure 2

Analysis of IFNG mRNA expression in T cells using manual and LWA semiautomated FISH-Flow protocols. PBMCs from an LTBI⁺ donor were either not stimulated or were stimulated with purified protein derivative (PPD) or staphylococcal enterotoxin B (SEB) for 6 h. Top row: cells were incubated in a 24-well plate and processed according to the PROCEDURE. Bottom row: cells were incubated in polystyrene FACS tubes and processed according to the LWA semiautomated protocol (Box 1). Frequencies of cells expressing IFNG mRNA in CD3⁺ and CD3⁻ cells are reported. Gates were set based on FMO controls and unstimulated samples. Cy5 fluorescence was recorded using a 640-nm laser and a 670/30 BP filter. A total of 100,000 events per sample were analyzed.

Figure 3

Flow cytometric analysis of Irga6 mRNA and IRGA6 protein in mouse macrophage cell line RAW264.7. Cells were stimulated with 100 U ml⁻¹ of IFN-γ or left unstimulated and collected at the indicated time points. Samples were processed according to the PROCEDURE. Cy5 fluorescence was recorded using a 640-nm laser and a 670/30 BP filter, whereas Alexa 405 fluorescence was recorded using a 405-nm laser and a 450/50 BP filter. A total of 100,000 events per sample were analyzed. (a) Kinetics of Irga6 mRNA and IRGA6 protein induction by IFNG-γ stimulation of RAW264.7 cells. (b) Scatter plots showing peak production of Irga6 mRNA (after 9 h of stimulation, upper panels) and IRGA6 protein (after 24 h of stimulation, lower panels). The frequencies of cells positive for Irga6 mRNA and IRGA6 protein are reported above each gate. Gates were set based on unstimulated samples.

Figure 4

Installation of the external tank in the LWA instrument. (a) Tubes with a three-way Luer-lock stopcock for connecting the instrument with the external tank. (b) Removal of the tube from the LWA wash tank and the barb connection on the instrument bulkhead. (c) Attachment of the bulkhead connector tube to the bulkhead cell wash connector. (d) Connection of the shorter tube to the internal cell wash tank (containing PBS-FISH buffer). (e) Connection of the longer tube to the external cell wash tank (HWB). (f,g) Positioning of (f) the stopcock and (g) the tubing to avoid tube tightening or pinching. The level sensor in panels b and c was removed to illustrate the barb connection on the instrument bulkhead.

Figure 5

Multiplex five-color analysis of IFNG mRNA expression in PBMC subsets. PBMCs from an LTBI⁺ donor were left unstimulated, or were stimulated with PPD for 6 h or with PMA/ionomycin for 2 h, and then were processed using the LWA semiautomated protocol (Box 1). In addition to hybridizing the cells with Cy5-labeled probes against IFNG mRNA, they were also stained for CD3, CD4, CD8 and CD56 surface markers using antibodies labeled with the indicated dyes. (a) Gating strategy. Gates on each cell population are indicated by gray boundaries. From left to right: single-cell gating includes single cells and excludes aggregates using FSC-A and FSC-H; lymphocytes are gated using FSC-A and SSC-A; CD3⁻ and CD3⁺ cells are gated using SSC-A and the FITC channel; and CD4⁺ and CD8⁺ cells are gated using the BUV395 and BUV496 channels on the CD3⁺ population. Gates for CD3, CD4 and CD8 were set using FMO controls. (b) Frequencies of IFNG⁺CD4⁺ and IFNG⁺CD8⁺ cells under unstimulated conditions, under PPD stimulation and under PMA/ionomycin stimulation. A lower total number of cells were acquired for the PMA/ionomycin stimulation relative to the other two conditions, because that stimulation leads to significant cell loss. Gates were set based on the unstimulated samples. (c) Relative frequencies of IFNG mRNA and CD56 expression in CD3⁻ cells. Gates were set based on FMO control and unstimulated samples. An example of the FMO control for staining without anti-CD56 antibody is shown in the panel (left). The following lasers and BP filters were used for each fluorochrome: Cy5, 640-nm laser and 670/30 BP filter; FITC, 488-nm laser and 530/30 BP filter; BUV395 and BUV496, 355-nm laser and 379/28 and 513/30 BP filters, respectively; PE-CF594, 561-nm laser and 610/20 BP filter. A total of 200,000 events per sample were analyzed. PMA, phorbol 12-myristate 13-acetate.