INTRODUCTION: Microarray technologies have made it possible to perform high-throughput and comprehensive analyses of nucleic acid sequences and expression. However, the technology to efficiently obtain high yields of RNA and DNA suitable for array analysis from purified populations of neoplastic cells from human tissues has not been well addressed. Microdissection can enrich populations of cells present in various tumour tissues, however it is not easily automated for high throughput use. In addition, there are some tissues in which it is difficult to separate neoplastic from contaminating normal cells. We have previously shown that DNA suitable for high throughput genotyping can be obtained from nuclei of neoplastic cell populations purified by flow cytometric cell sorting. To apply array technology to in vivo studies of human neoplastic progression, we have developed protocols for rapid and efficient isolation of RNA and DNA from flow-cytometrically purified whole epithelial cells from fresh and frozen primary tissue. We have also utilised a real-time PCR assay to determine RNA quality after flow sorting. MATERIALS AND METHODS: A549 lung adenocarcinoma cells were cultured in RPMI supplemented with 5% FCS. Cells were trypsinised according to standard protocols, resuspended as single cells in the aqueous reagent RNAlater™ (Ambion, Inc. Austin TX) and kept on ice. Breast tissues were obtained from mastectomy samples and biopsies frozen in minimal essential media (Gibco, Rockville MD) containing 10% dimethyl sulfoxide (Fisher, Fair Lawn, NJ). Single-cell suspensions were created by placing fresh or thawed tissues in phosphate buffered saline with 1% bovine serum albumin (BSA; PBA), and mincing. The minced tissue was resuspended in PBA, filtered through 40 u mesh, centrifuged then resuspended in 100 ul of RNAlater™. Single-cell suspensions were stained with epithelial-specific cytokeratin CAM 5.2-FITC (Becton Dickinson, San Jose, CA) or a FITC-labeled isotype matched mouse immunoglobulin IgG2a-FITC (Becton Dickinson, San Jose, CA) on ice for 1 hour. Samples were diluted with 1.5 ml of PBA then resuspended in 0.5–1.0 ml of 10 g/ml 2,4 diamidino-6-phenylindole (DAPI) in NST buffer (146 mM NaCl, 10 mM Tris base, 0.1% NP40 (Sigma), 2 mM Ca, 20 mM Mg, and 0.05% BSA (Sigma), pH 7.4). All centrifugation steps were performed at 4×g for 8 minutes at 2°C and all reagents were DNAse/RNAse free. Whole cells were sorted by DNA content and cytokeratin staining intensity on a Becton Dickinson Vantage flow cytometer, with 100 mW, 488 nm excitation (FITC and PI) and 100 mW, 354–361 nm (UV) excitation (DAPI). The instrument was aligned daily using “DNA check” beads (Coulter, Miami, Fl.) to give coefficients of variation of <3%. List-mode data were analysed using MultiCycle AV and MultiPlus AV software (Phoenix Flow Systems, San Diego, CA). RNA was isolated according to standard protocols, while DNA was simultaneously extracted from organic phases. The integrity of RNA from flow sorted cells was determined by real-time PCR assays, TaqMan® EZ RT-PCR (PE Applied Biosystems, Foster City, CA) using primers and probes specific for the 5′ and the 3′ regions of GAPDH. RESULTS. A549 and primary breast epithelial cells fixed in RNAlater™ showed good staining patterns with DNA dyes and CAM 5.2 cytokeratin antibodies. Although fixation was effective after 1 hour in the reagent, cytokeratin fluorescence intensity was further enhanced and DNA content CVs were improved by continued incubation of cells for longer periods of time. Visual observation by fluorescence microscopy showed that epithelial cells were intact after sorting at a rate of 2–3×103 events/sec. Equivalent amounts of total RNA (3–10 µg/106 cells) were isolated from fixed sorted and fixed unsorted epithelial cells. The ratios of real-time PCR threshold concentrations (CT) for the GAPDH probes (3′/5′ CT) were equivalent for control RNA, and for RNA from unsorted and sorted epithelial cells fixed and stained in RNAlater™. Cells fixed in EtOH/acetic acid had poorer mRNA yields and lower CT ratios. Furthermore, whereas scrupulous cleaning of the cytometer tubing and use of the RNAse inhibitor DEPC in sheath fluid was required in order to preserve message in EtOH/acetic acid fixed cells; such treatment was not required for cells fixed in RNAlater™. Agarose gel analysis showed that high CT values correlated with intact 28S and 18S rRNA bands. DNA extracted simultaneously with RNA from sorted epithelial cells provided a template for whole genome amplification, genotyping and sequencing. CONCLUSIONS. The aqueous reagent, RNAlater™, which fixes and preserves RNA, allows staining and purification of whole epithelial cells by flow sorting. High yields of RNA and DNA suitable for expression and genotype analysis can be obtained from flow cytometrically purified populations of neoplastic cells from tissues in vivo.
This is a preview of subscription content, access via your institution