LINCS MCF 10A Common Project: Rolling-time-point sensitivity measures of the MCF 10A breast cell line to 8 small molecule perturbagens. Dataset 15 of 15: Time-dependent dose-response metrics for biological replicate 3. - Dataset (ID:20323)
HMS Dataset ID: | 20323 |
Dataset Title: | LINCS MCF 10A Common Project: Rolling-time-point sensitivity measures of the MCF 10A breast cell line to 8 small molecule perturbagens. Dataset 15 of 15: Time-dependent dose-response metrics for biological replicate 3. |
Screening Lab Investigator: | Mario Niepel, Marc Hafner |
Screening Principal Investigator: | Peter K. Sorger |
Assay Description: | To generate measures of the sensitivity of a derivative of the MCF 10A cell line to 8 small molecule perturbagens, we treated cells with single drugs over a minimum 11-point dilution series centered around the GR50 but not exceeding 10 µM and then measured cell number continuously for up to 90 hours after drug exposure. |
Assay Protocol: |
1. Cells in mid-log phase of the growth cycle for an H2B-mCherry-expressing derivative of MCF 10A were plated at 750 cells/well in 60 µL of complete growth medium (DMEM: F-12 + 5% (v/v) horse serum + 10 µg/mL human insulin + 20 ng/mL rhEGF + 100 ng/mL Cholera toxin + 0.5 µg/mL Hydrocortisone) in duplicate 384-well plates. 2. The plated cells were grown for 24 hours at 37°C in the presence of 5% CO2. 3. The plates were treated with DMSO or with the indicated doses of small molecule by direct dispensing of DMSO stock solutions to the indicated concentrations into 60 µL of media using an HP D300 Digital Dispenser. At the same time, YOYO-1 Iodide (Thermo Fisher Scientific) was added to each well to a concentration of 250 nM. 4. The cells then were imaged every 2-4 hr for up to ~90 hr using an Operetta (Perkin Elmer) high-content imaging system equipped with a live-cell chamber. 5. To obtain cell counts at each time point, the number of total nuclei and, of that total, the numbers of live and dead cells were quantified using the Columbus image data storage and analysis system. Total nuclei counts were obtained using H2B-mCherry fluorescence, and dead cell counts were obtained using YOYO-1 Iodide staining. 6. For each time point, cell counts were normalized to DMSO-treated controls on the same plate to yield relative cell count and normalized growth rate inhibition (GR) values for each time point, plate (technical replicate), and drug/drug concentration combination (HMS LINCS Datasets #20309-20311). Consistent with the methods reported in Hafner et al. (2016) (PMID: 27135972), relative cell count x(c,t)/xctrl(t) is the measured cell count x(c,t) for a given treatment at time t divided by xctrl(t), the 50%-trimmed mean of the cell count of the DMSO-treated control wells on the same plate at the same time t. GR values were calculated according to the following formula: 2^[log2(x(c,t)/x0)/log2(xctrl/x0)]-1 where x(c,t) is the measured cell count after a given treatment at time t, xctrl(t) is the 50%-trimmed mean of the cell count of the DMSO-treated control wells on the same treated plate at the same time t, and x0 is the 50%-trimmed mean of the cell count at time 0 on the same plate. Data for three complete biological replicates were analyzed. 7. Within each experiment, the results of two plates (technical replicates) for each time point were averaged to yield the mean relative cell count and the mean GR value for each drug/drug concentration combination for a given biological replicate at each time point (HMS LINCS Datasets #20312-20314). 8. For each small molecule and at each time point, mean GR values across all tested concentrations were fitted to a sigmoidal curve to extract the GR50, GEC50, GRmax, GRinf, Hill coefficient, and GRAOC. The values from measurements made every ~12 hr are reported (HMS LINCS Datasets #20315-20317). 9. In addition, time-dependent GR metrics were calculated using an ~12-hr rolling-window approach. The difference in nuclei counts between two selected time points for a given treated well was normalized to the difference for the DMSO-treated controls on the same plate to yield time-dependent GR values for each plate (technical replicate) and drug/drug concentration combination. Time-dependent GR values were calculated according to the following formula: 2^[log2{x(c,t+Δt)/x(c,t-Δt)}/log2{xctrl(t+Δt)/xctrl(t-Δt)}]-1 where x(c,t±Δt) are the measured cell counts after a given treatment at times t+Δt and t-Δt and where xctrl(t±Δt) are the 50%-trimmed means of the cell counts of the DMSO-treated control wells on the same treated plate at the same times t+Δt and t-Δt. The time interval 2×Δt is reported in each dataset. 10. Within each experiment, the results of the two plates (technical replicates) were averaged to yield the mean time-dependent GR value for each drug/drug concentration combination for a given biological replicate for each time interval (HMS LINCS Datasets #20318-20320). Note that the two plates for a given biological replicate were not scanned at exactly the same time points, so the time point and time interval reported in the dataset correspond to the average of the time points for the two replicates used to calculate the mean. 11. For each small molecule and at each time point, time-dependent mean GR values across all tested concentrations were fitted to a sigmoidal curve to extract the time-dependent GR50, GEC50, GRmax, GRinf, Hill coefficient, and GRAOC for the specified time points and time intervals (HMS LINCS Datasets #20321-20323). |
Assay Protocol Reference: | Hafner, M., Niepel, M., Chung, M., and Sorger, P.K. (2016) Growth rate inhibition metrics correct for confounders in measuring sensitivity to cancer drugs. Nat Methods. 13(6):521-527. doi:10.1038/nmeth.3853 PMID: 27135972 PMCID: PMC4887336 |
HMS Dataset Type: | Analysis |
Date Publicly Available: | 2017-05-12 |
Most Recent Update: | 2017-05-12 |