Secondary siRNA screen (5-ethynyl uridine RNA metabolic labelling)
Secondary image-based genetic screen targeting 463 genes in human HeLa cells using RNA interference (siRNA), as described in Berry et al, 2021. Cells incubated with 5-ethynyl uridine (5-EU) for 30min before fixation. Nascent RNA detected by click reaction with AlexaFluor488-azide. Other stains: DAPI (DNA), proliferating cell nuclear antigen (PCNA), succinimidyl ester AlexaFluor647 (total protein). Cells fixed 72h after siRNA transfection. Imaged at 20X magnification with two experimental replicates per condition. Further description available in Berry et al., 2022. Single-cell features provided as background-subtracted mean and sum intensity values for the nucleus, cytoplasm, or whole-cell, as appropriate, for each fluorescence channel. Cell-cycle (G1/S/G2) annotation using a classifier trained using PCNA and DAPI features on EdU-incubated cells, as described in the manuscript. Data cleanup and correction applied as described in the manuscript. Summaries provided on a per-well basis.
Steps to reproduce
900 HeLa cells were plated per well in 384-well plates for reverse transfection onto a mixture of pooled siRNAs dispensed using an acoustic dispenser and Lipofectamine RNAiMAX (0.08µl per well in 10µL OptiMEM) according to manufacturer’s specifications. Cells were subsequently grown for 72 hours at 37°C in a final volume of 40µL growth media, to establish efficient knockdown of the targeted genes. Cells were fixed for 15 min in 4% paraformaldehyde, permeabilised in 0.2% Triton X-100 for 15 min. Nascent RNA was visualised using metabolic labelling as previously described (Jao et al., PNAS, 2008), with modifications. Briefly, adherent cells were cultured in complete media at 37°C, 5% CO2 for 2-3 days. 5-ethynyl uridine (EU) was then dissolved to a concentration of 2mM in pre-warmed complete media. EU was added to cells by partially aspirating growth media and dispensing an equal volume of 2mM EU using a BioTek washer-dispenser (e.g. 30µL 2mM EU added to 30µL residual for 384-well plates: final EU concentration = 1mM). Cells were then incubated for 20 or 30 min at 37°C, 5% CO2, before fixation with 4% PFA at room temperature for 20-30 min. After fixation, cells were permeabilised with 0.5% Triton X-100 and washed 3 times with TBS (50mM Tris pH 8.0, 150mM NaCl). To render nascent RNA fluorescent, we prepared sufficient volume of click reaction master mix for all wells on a plate at 1.5x concentration, as follows: 75µM Alexa Fluor 488 azide or Alexa Fluor 647 azide, 3mM CuSO4, 150mM Sodium ascorbate, in TBS. Click reaction was dispensed using a BioTek washer-dispenser and incubated for 30 min at room temperature before washing cells 3x into PBS. Samples were imaged on an automated spinning-disk microscope (CellVoyager 7000, Yokogawa), which is equipped with four excitation lasers (405, 488, 568, 647nm) and two Neo sCMOS cameras (Andor). A 20X/NA0.75 objective was used. Computational image analysis was used to segment cells and nuclei from max-projected images, and intensity features calculated for all channels in the nucleus, cytoplasm and cell. Mitotic/apoptotic cells were identified and removed using iterative supervised machine learning in tissuemaps. Intensity values were background-subtracted and quadrant effects arising from pipetting inconsistencies were corrected using median normalisation. Intensity values were further median normalized per plate using negative control wells on each plate (scrambled siRNA).