Fig 1.
Schematic representation of formaldehyde and glyoxal reactions with guanine nucleotides.
A: Reaction of formaldehyde with guanine, equivalent reactions occur for adenine and cytidine. B: Reaction of glyoxal with guanine, reaction products with other bases are not stable [34].
Fig 2.
Determination of RNA-compatible fixation and permeabilisation conditions.
A: 1x106 COLO205 cells were either unfixed (lane 1), fixed with 70% ethanol on ice for 15 minutes (lane 2) or fixed with 4% formaldehyde on ice for 15 minutes (lane 3). Unfixed cells were dissolved immediately in TRI Reagent, fixed cells were washed once in PBS by centrifugation at 2000 x g for 3 minutes at 4°C before RNA extraction with TRI Reagent. 20% of RNA obtained was separated on a 1.2% glyoxal gel and imaged by ethidium bromide staining. B: COLO205 cells were either unfixed (lane 1), fixed with 70% ethanol on ice for 15 minutes (lane 2) or fixed with 100% methanol on ice for 15 minutes (lane 3). Unfixed cells were dissolved immediately in TRI Reagent, fixed cells were washed once in PBS by centrifugation at 2000 x g for 3 minutes at 4°C before RNA extraction with TRI Reagent. RNA was analysed as in A. C: COLO205 cells were fixed with glyoxal fixation mix (pH4) either without or with 20% ethanol and incubated on ice for 15 minutes and washed once in PBS by centrifugation at 2000 x g for 3 minutes at 4°C. RNA was extracted and analysed as in A. D: COLO205 cells were either unfixed (lane 1), fixed with glyoxal fixation mix (pH4) with 20% ethanol (lane 2) or with 4% formaldehyde on ice for 15 minutes (lane 3). Cells were washed once in PBS by centrifugation at 2000 x g for 3 minutes at 4°C and incubated on ice for 1 hour in 100 μl PBS followed by centrifugation at 2000 x g for 3 minutes at 4°C before RNA extraction with an RNeasy mini kit. E: 100 ng RNA per reaction from D was subjected to one-step combined reverse transcription and quantitative PCR reactions for ACT1B, GAPDH and PGK1. The RNA yield from formaldehyde-fixed RNA cells was <<100ng and therefore control and formaldehyde-fixed reactions were performed on cell equivalents of extracted RNA solution. Ct is the cycle number at which the fluorescence exceeded threshold. 3 technical replicates for each RT-qPCR reaction were performed. F: COLO205 cells were either unfixed (lane 1) or fixed with glyoxal fixation mix (pH4) with 20% ethanol on ice for 15 minutes and permeabilised in 100% methanol on ice for 30 minutes (lane 2), or 0.5% saponin on ice for 30 minutes (lane 3) or 0.3% Triton X-100 on ice for 30 minutes (lane 4). RNA was analysed as in A. G: COLO205 cells were either unfixed (lane 1) or fixed with glyoxal fixation mix (pH4) with 20% ethanol on ice for 15 minutes, permeabilised in 100% methanol on ice for 30 minutes followed by incubation in primary antibody for 1 hour on ice and secondary antibody for 30 minutes on ice in dark (lane 2). RNA was analysed as in A.
Fig 3.
RNA-seq of stained and sorted cell populations.
A: Bioanalyzer profiles of total RNA derived from unprocessed cells (harvested directly from trypsinised cells) compared to cells that have undergone glyoxal fixation, permeabilisation with 100% methanol and indirect immunofluorescence for CCNB1. 0.1–5 ng of total RNA was separated on total RNA pico chips on Agilent 2100 Bioanalyzer. B: Scatter plots comparing the normalised read count for each annotated gene in GRCh38 in poly(A)+ RNA-seq libraries derived from unprocessed and from glyoxal fixed, permeabilised and stained cells. Two biological replicates were sequenced for each condition and averaged, read counts were normalised to the total number of reads in each library. Data is shown for both COLO205 and MCF-7 cells. C: Flow cytometry density plots for MCF-7 cells labelled with anti-CCNB1 primary antibody and donkey Alexa Fluor-488 conjugated secondary antibody and sorted using a BD FACSAria III sorter. Intact cells (shown within the elliptical gate) were distinguished from off-scale events and cell debris using forward scatter (FSC) and side scatter (SSC) measurements (panel 1). Doublets were excluded from the gated cells by a 2-step gating strategy with pulse height (H) plotted against area (A) for FSC parameter followed by SSC-H versus SSC-A plot (panel 3). Fluorescence thresholds for isolation of CCNB1 positive and negative cell fractions are shown in panel 4. Gates were set using the negative control and the CCNB1 positive and negative sorting gates were moved apart from each other to maximise purity when sorted. D: Fluorescence histogram of DAPI intensities for sorted CCNB1 negative and positive populations. E: Scatter plots comparing the normalised read count for each annotated gene in GRCh38 in poly(A)+ RNA-seq libraries derived from two biological replicates each of CCNB1 positive (left) and negative (right) MCF-7 cells. Note the particularly tight correlation after sorting towards the antigen of interest. E: Hierarchical clustering analysis of the 161 significantly different expressed genes identified DEseq2 analysis (p<0.05) comparing CCNB1-positive and CCNB1-negative MCF-7 samples.
Table 1.
Details of RNA extraction and library construction.