Figure 1.
CMV and EF1α promoter comparison in stable cell lines.
MCF10A cells expressing human ORF constructs were selected by flow cytometry to minimum 98% GFP positive, seeded and fixed for analysis on indicated day following sorting. A,B,C: Bar graphs representing the proportion of GFP (A,C) or EdU (B) cells obtained by high-content imaging of cells in 96 well plates. A, B: CMV promoter constructs; C: EF1α promoter constructs. All bars represent mean of 4 wells (error bars = SD). D: frequency histograms of cells analysed for GFP fluorescence by flow cytometry, control data (black curves) are for the parental, untransduced cells, black bars indicate threshold used for selecting GFP positive cells during sorting.
Figure 2.
Diagram summarising the key library construction steps.
1. Gateway cloning was used to transfer ORFs into the lentiviral expression plasmid plv411G downstream of the EF1α promoter and upstream of the IRES driven GFP; 2. Virus production was performed by transfecting expression clones together with viral packaging plasmids into HEK293T cell line; 3. Viral supernatant was collected into a fresh set of plates and stored; 4. Transfected cells remaining in the plates were fixed and scanned to determine GFP fluorescence in each well. Frequency distribution histogram illustrates number of wells (y-axis) that had similar GFP fluorescence intensity (x-axis). Values between plates were normalized by zeroing on the mean of 4 mock transfected wells on each plate. Each plate also contained 4 empty expression vector wells, shown in green, which were used as positive controls; 5. Thirty eight ORF-expressing wells were randomly selected from each of the tail ends of the frequency distribution categories, to evaluate the performance of the viral supernatants on a range of target cell lines.
Figure 3.
Library clone insert size and GFP fluorescence in virus-making cells.
A: scatter-plot of cloned insert size and total well GFP fluorescence in transfected HEK293T cells. B: frequency distribution histogram of clones in different insert size category.
Figure 4.
A- Pseudo-coloured overlay (MERGE: GFP channel green, DAPI channel, blue, bar = 50 µm) of images from one scanned field within a well and object analysis masks for the enlarged boxed area in DAPI and GFP channels. DAPI channel was used to determine nuclear area mask (blue outline). Nuclear area was then analysed for average pixel intensity in the GFP channel, and threshold set visually to distinguish GFP positive (green outline) from GFP negative nuclei (orange outline). B-Average nuclear GFP pixel intensity frequency histograms for wells representing untransduced (mock), or cells transduced with low (SCAF8) or high (KCTD2) titer gene supernatant. Mock well histograms were used to validate and/or adjust intensity threshold (red line) for identifying GFP positive cells. Images and data are from HaCat cell line, taken with a 10x objective, image scanning and analysis was performed using Cellomics TargetActivation v3. application algorithm.
Figure 5.
Transduction rates observed with low and high titer virus-producing expression clones.
A,C – low titer clones; B,D –high titer clones. The bars in the inset in graph A represent mean and standard deviation (error bars) of four wells transduced with empty vector virus. Color of bars corresponds to the target cell line as indicated in the inset graph. Actively dividing human epithelial tumor (MCF7, PMC42-ET, MDA-MB-468, WMM1175), and non-tumor (MCF10A, HaCaT) cell lines were compared to non-dividing cells (WMM1175-p16, arrested by induced overexpression of p16, and primary mouse bone marrow macrophages (mBMM)). The proportion of GFP positive cells (y-axis) was determined after high-content imaging of plates containing fixed transduced cells. Each bar in A and B represents data for a well transduced with a single gene-expressing vector as indicated on x-axis. Bars representing mean values for vector wells are included in all three graphs to allow for scale comparison. Each bar in C and D represents a mean (error bar = SD) transduction rate for all human cell lines for a given gene.
Figure 6.
Comparison of target cell transduction rate and GFP fluorescence of corresponding virus-producing wells.
A,C – target cell transduction rates; B,D – GFP fluorescence of virus producing wells. Bars in A and B represent mean of 4 wells derived from independent single colony isolates of the gene-expressing plasmids (x-axis). Genes underlined in A, produced high viral titer in bulk well experiments (Figures 1 and 2.), the rest were low titer producers. Error bars = SD. Data for graphs A and C were obtained by high-content imaging of transduced cells, while data for graphs B and D were obtained by scanning the transfected cells using a fluorescence plate reader. C and D: Mean and SD for high and low titer well data represented in A and B. In all cell lines mean for combined low titer genes was significantly different from the mean for high titer wells (Aspin-Welch test, P value: HEK293T = 4.70 E−11; HACAT = 5.68 E−23; MDA-MB468 = 2.32 E−36; WMM1175 = 7.29 E−12; MCF10A = 2.98 E−21).
Table 1.
Genes producing very high or very low viral titer.
Figure 7.
Effect of low virus titer-producing genes on viability of the transfected cells.
HEK293T cells were transfected with gene (x-axis) expressing plasmids either alone or with viral packaging vector day after seeding. 21 h after transfection, medium was replaced with PBS containing Hoechst33342 and propidium iodide (PI), which stain DNA and dying cells respectively. Cells were scanned in blue (Hoechst, A) red (PI, B) and green (GFP, C) channels and number of objects determined in each channel. A and B represent independent object counts, data in C are expressed as proportion of GFP positive cells in the Hoechst stained population. A - Significantly reduced number of surviving cells in the well compared to untreated wells (CELLS): lipofectamine alone (MOCK) (P = 0.034); MOCK with packaging plasmids (P = 7.5E−10); all others (P≤1.1E−05) B - Significantly increased PI positive cells compared to vector: expression plasmid only BCL2L1 (P = 6.88E−05), C3ORF1(P = 0.03), MTCH1(P = 4.66E−05), PNMAL2 (P = 0.01) P values determined using Aspin-Welch test; Bars = mean of 4 wells, error bars = SD, CELLS-untreated well, MOCK-wells where expression plasmid has been omitted. AWAT2, and vector were high virus titer producing in previous experiments.