Julfa Begum and William Day contributed equally to this manuscript.
A method for evaluating the use of fluorescent dyes to track proliferation in cell lines by dye dilution
Article first published online: 25 OCT 2013
© 2013 International Society for Advancement of Cytometry
Cytometry Part A
Volume 83, Issue 12, pages 1085–1095, December 2013
How to Cite
Begum, J., Day, W., Henderson, C., Purewal, S., Cerveira, J., Summers, H., Rees, P., Davies, D. and Filby, A. (2013), A method for evaluating the use of fluorescent dyes to track proliferation in cell lines by dye dilution. Cytometry, 83: 1085–1095. doi: 10.1002/cyto.a.22403
- Issue published online: 22 NOV 2013
- Article first published online: 25 OCT 2013
- Cancer Research UK
- EPSRC. Grant Number: EP/J00619X/1
Additional Supporting Information may be found in the online version of this article.
|cytoa22403-sup-0001-suppfig1.tiff||1405K||Supplementary Information Figure 1: Relationship between dye concentration and (A) staining intensity (MedFI) (B) uniformity (CV) post-labelling and (C) cell viability (PI exclusion) of Jurkat cells after 24 hrs. (D) Bi-Variate plots showing the toxicity of each dye at 48 hrs. for the indicated concentrations. (E) Plots of Jurkat cells stained with the optimal dye concentrations and spiked with 6 peak beads before acquisition on an LSR Fortessa system. Please note that the PMT settings are not optimal for 6-peak resolution, rather they are set so as to try and place the beads and dye labelled cells on scale. This is to provide context to the level of fluorescence achieved for each dye labelling and therefore the potential influences that will affect population spreading around the mean.|
|cytoa22403-sup-0002-suppfig2.tiff||1129K||Supplementary Information Figure 2: Sorting and re-analysing channel-matched fluorescent beads provides an estimation of detector-specific extrinsic errors and cell/dye intrinsic errors. Panels of individual bead sets ranging from 3 - 10 μm, containing dyes with either identical or analogous spectral properties to CSFE (A), CTV (B) or EPD (C) were sorted with increasing channel widths and re-analysed. Each graph shows width of the post-sorted populations with the inclusions of the respective dye labelled Jurkat cells for reference. (D) A graph showing the contribution extrinsic and intrinsic error sources make to the post-sort channel widths shown in figure 1B. Extrinsic error was determined for the detector channel of each dye using fluorescence beads with matched spectral properties to the dyes under investigation (Fig S2A-C) and an analogous sort strategy (dotted lines). These values were then subtracted from values in 1B using equation 2 to determine the dye-specific contribution to post-sort channel width increase (solid lines). In all cases, the mean +/- SEM of 4 independent experiments is shown. (E) 6-peaks Cyto-Cal bead data showing our measure of performance of the respective detector channels run on the Aria. The PMT voltages are shown on the top left of each plot. (F) The contribution that cell intrinsic error makes to the re-spreading values shown in figure 3B by correction using the values shown in figure 3C by equation 2. The dotted line denotes the threshold for peak resolution based on the cell-intrinsic width from on empirical data shown in figure 1C.|
|cytoa22403-sup-0003-suppfig3.tiff||2165K||Supplementary Information Figure 3: The analysis strategy for determining dye transfer and apportionment at telophase using IFC/FlowSight. (A) The gating and analysis strategy for identifying single cells for measuring dye transfer under pair-wise culture conditions. Single cells were selected using the aspect ratio and area of the BF channel mask (1). Focused cells selected by the gradient RMS score for the BF image (2) and total intensity of the respective channels was plotted (3). (B) The gating and analysis strategy for identifying single and bi-lobular MPM2+ 4N cells and subdividing into the four major mitotic stages by nuclear morphometry. Again, BF area and Aspect ratio were used to identify single and bi-lobular cells (1). MPM2+ 4N cells were identified from within the single cell gate (2). The bi-lobular gate (3) and the OR Boolean population of both (2+3). The adapted nuclear mask shown in C was then used to calculate and plot the nuclear spot count (x) and aspect ratio (y) of the MPM2+ single and bi-lobular cells to gate on Prophase, Metaphase and Ana/Telophase (4). Anaphase and telophase cells were subdivided using the Area (x) and aspect ratio (y) of the MPM2 channel image (5). (C) The masking and gating strategy for identifying the 4 main mitotic phases, the IDEAS language string is shown below the images (D) The image analysis strategy for measuring the distribution of CFSE, CTV, EPD and PI across the putative daughter poles of telophasic cells using exported 16-bit TIFF images of validated telophasic cells and Image J.|
|cytoa22403-sup-0004-suppfig4.tiff||1491K||Supplementary Information Figure 4: A549 cells were labelled with the indicated titration of CSFE, CTV and EPD followed by acquisition on an LSRFortessa, with the % viable by PI exclusion (A), MedFI (B) and CV (C) all plotted. (D) A549 cells were acquired on a FlowSight and representative images (20x) are shown of the individual dye staining patterns, again note the punctate EPD staining as well as the (E) transfer of EPD to CTV-labelled cells under co-culture conditions (black dots) compared to each population labelled alone (grey dots). (F) Proliferation traces of CTV-labelled Jurkats or A549 cells sorted on the influx or Aria, acquired on an LSRFortessa. Sorted cells were placed into culture for 48 hrs. and then analysed for division peak resolution. The cell intrinsic errors from equation 1 are shown in the top left of each plots for a given sort gate and relate to the widths shown in figure 3B. NA = Not applicable.|
|cytoa22403-sup-0005-suppfig5.tiff||1061K||Supplementary Information Figure 5: The differences in proliferative capacity of lCTV and hCTV cells can be explained by the fact cells in G2/M stain brighter for the tracking dyes than those in G1/S. (A) A graph of the median FSC-A measurements of the unsorted, hCTV and lCTV populations (Aria). (B) Bi-variate plots (left panels) of CFSE, CTV and EPD labelled cells fixed in ethanol and stained for PI (x-axis) and pH3 AF488 (y-axis) analysed by LSRFortessa. The gated frequencies are shown. Histogram overlays of the respective dye intensity for each gated cell cycle phase are shown (right panels) with the median values included. (C) A graph showing the serum dependency for driving proliferation of hCTV and lCTV cells in culture (B and C analysed on an LSRFortessa). (D) The sort decision for eliminating G2/M cells from the population prior to selecting hCTV and lCTV cells (Aria).|
|cytoa22403-sup-0006-suppfig6.tiff||739K||Supplementary Information Figure 6: Attempts to limit biological heterogeneity prior to CTV labelling did not sufficiently reduce width enough to resolve division peaks. (A) Jurkat cells were cultured for 32 hrs. in the presence or absence of 10% FCS after which cell were labelled with CTV. PI stained cell cycle profiles from EtOH fixed aliquots of the same cells to determine the success of synchronisation with a gate set on the G2-M population. The 8 bit scaled channel width of the CTV signal is shown in the upper left of the CTV histograms. (B) Jurkat cells were sorted using incrementally increasing sort channel widths set on the SSC and labelled with CTV (upper panels), these cells were then cultured for 48 hr. to determine division peak resolution.|
Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.