stke.sciencemag.org/cgi/content/full/13/636/eaay1451/DC1

Supplementary Materials for

High-throughput dynamic BH3 profiling may quickly and accurately predict

effective therapies in solid tumors

Patrick Bhola, Eman Ahmed, Jennifer L. Guerriero, Ewa Sicinska, Emily Su, Elizaveta Lavrova, Jing Ni, Otari Chipashvili, Timothy Hagan, Marissa S. Pioso, Kelley McQueeney, Kimmie Ng, Andrew J. Aguirre, James M. Cleary,

David Cocozziello, Alaba Sotayo, Jeremy Ryan, Jean J. Zhao, Anthony Letai*

*Corresponding author. Email: anthony_letai@dfci.harvard.edu

Published 16 June 2020, Sci. Signal. 13, eaay1451 (2020)

DOI: 10.1126/scisignal.aay1451

The PDF file includes:

Text S1. Detailed HT-DBP protocol. Fig. S1. Schematic of DBP. Fig. S2. Selection of informative peptide screening concentrations. Fig. S3. Identification of a drug screening concentration. Fig. S4. HT-DBP enables identification of compounds that sensitize cancer cells for apoptosis. Fig. S5. Similarity between FACS and microscopy dynamic BH3 profiles. Fig. S6. Images of stained MMTV-PyMT tumor cells. Fig. S7. Example of cell masks. Fig. S8. Screening data for freshly isolated MMTV-PyMT tumors. Fig. S9. Counterscreens in healthy cells. Fig. S10. Supplemental data for in vivo validation experiments for HT-DBP. Fig. S11. Identification of compounds that sensitize the COCA9 colon cancer PDX for apoptosis. Fig. S12. Quantification of apoptotic chemical vulnerabilities in colon cancer PDX models. Fig. S13. Heat map of apoptotic chemical vulnerabilities in colon cancer PDX models. Fig. S14. Comparison of apoptotic chemical vulnerabilities in colon cancer PDX models. Fig. S15. Supplemental data for comparison of chemical vulnerabilities of freshly isolated and cultured cancer cells. Fig. S16. Supplemental data for comparison of chemical vulnerabilities of freshly isolated and cultured cancer cells. Legend for table S1 Legends for data files S1 to S16

Other Supplementary Material for this manuscript includes the following: (available at stke.sciencemag.org/cgi/content/full/13/636/eaay1451/DC1)

Table S1 (Microsoft Excel format). List of compounds used in the chemical screen. Data file S1 (Microsoft Excel format). Cell count and delta priming of MDA-MB-231 line. Data file S2 (Microsoft Excel format). Apoptosis and delta priming of MDA-MB-231 line. Data file S3 (Microsoft Excel format). Cell count and delta priming of MMTV-PyMT tumor. Data file S4 (Microsoft Excel format). Technical replicate of MMTV-PyMT delta priming. Data file S5 (Microsoft Excel format). Biological replicate of MMTV-PyMT delta priming. Data file S6 (Microsoft Excel format). Cell count and delta priming of primary mouse hepatocytes. Data file S7 (Microsoft Excel format). Cell count and delta priming of human foreskin fibroblasts. Data file S8 (Microsoft Excel format). Delta priming in adult mouse hepatocytes and MMTV-PyMT tumors. Data file S9 (Microsoft Excel format). Delta priming of freshly isolated MMTV-PyMT tumors by nominal drug target. Data file S10 (Microsoft Excel format). Normalized delta priming of colorectal PDX models. Data file S11 (Microsoft Excel format). Normalized delta priming of colorectal PDX models by nominal target. Data file S12 (Microsoft Excel format). Delta priming of MMTV-PyMT–derived cell line. Data file S13 (Microsoft Excel format). Correlation between MMTV-PyMT tumor and derived cell line. Data file S14 (Microsoft Excel format). Dose response by delta priming in MMTV-PyMT tumor and derived cell line. Data file S15 (Microsoft Excel format). Delta priming in colorectal PDX models. Data file S16 (Microsoft Excel format). Delta priming in primary human CRC cells ex vivo.

Text S1. Detailed HT-DBP protocol HT-DBP involves the steps below:

1. Cell plating and compound treatment. 2. Titration plate to identify the most informative BH3 peptide concentration. 3. Application of the ideal peptide concentration to compound treated plates. 4. Imaging the plates, and image analysis.

Cell plating and compound treatment: 30 µL of dissociated cells were plated in each well of a 384 well plate (Corning 3764 or 3542 plates) using the Thermo Combi multi-well dispenser (Thermo Fisher Scientific). Cell numbers per well varied from 1000-5000 cells per well depending on cell size and adhesive properties of cells. For mouse tumors or PDX models, cells were plated in collagen I coated 384 well plates. MMTV-PyMT cells were plated at a concentration of 4000 cells per well. DF-BM355 cells were plated at a concentration of 5000 cells per well. Colon PDX models were plated at a concentration of 4000-8000 cells per well. Plating for chemical screening was performed in technical duplicate. Cells were given time to settle to the bottom of plates for approximately 20-30 mins, and cells were subsequently drugged using pin transfer machines (ICCB at Harvard Medical School or Broad Institute). Cells and drugs (table S1) were incubated overnight at 37 °C in a 5% CO2 incubator. Titration plate: To identify the optimal peptide concentration for performing chemical screens, we performed a titration of the Bim BH3 peptide to identify the highest Bim peptide concentration where there was little or no loss of cytochrome c. We expect that this concentration will be the most informative at identifying compounds that sensitize cells for apoptosis. Bim BH3 titration plates were generated by making a master plate of 2X MEB solution with digitonin (to have 0.002% final for human cells, and 0.001% for mouse cells), and peptides. A linear, two-fold, 15-point titration was performed for all cells except for the BM355 tumors where there was limited number of cells, and where only an 8-point titration was performed. The titration was performed in at least technical triplicates. To perform the peptide titration, plated cells were washed three times in 1X PBS using the BioTek plate washer (using the “manifold wash” setting), leaving 15 µL residual PBS in the wells. A mixture from the 2X MEB master plate was added to each well (15 µL per well), and cells were incubated at 30 °C for 1 hour. Cells were fixed by adding 10 µL of 8% PFA for 15 mins. Plates were aspirated to a residual volume of 20 µL using the plate washer and 20 µL of the N2 neutralization buffer was added for 5 minutes. Plates were aspirated to a residual volume of 20 µL using the plate washer and 10 µL of a 3X staining solution was added. Cytochrome c was added at a concentration of 1:200. Hoechst was added to a concentration of 1:2000. Cells were incubated for 1 to 2 hours at room temperature then washed and imaged. Details on image analysis are provided below. Screening plates: Once the ideal peptide titration was calculated, a 2X MEB buffer was made with the appropriate concentration of the Bim BH3 peptide and digitonin. Cells were washed 3 times with 1X PBS to a residual volume of 15 µL. The 2X MEB mixture was added to cells (15 µL), and the cells were incubated for 1 hour at 30 °C. Cells were fixed by adding 10 µL of 8% PFA and waiting for 15 mins at room temperature. Liquid in wells were aspirated to 20 µL and 20µL of the N2 solution was added. Plates were incubated at room temperature for 5 mins. Liquid in wells were aspirated to 20 µL and 10 µL of the stain solution was added to cells. Staining solution consisted of 1:2000 of cytochrome c and Hoechst

33342 stain. Stain solutions were incubated overnight at 4C. Cells were subsequently washed 3 times in 1X PBS, and imaged. Image collection: Images were acquired using an IXM or IXM4 (Molecular Devices) microscope. Briefly, we used a 10x widefield objective to capture 1 to 4 images per well, representing about 60-80% of the well bottom, depending on the plate type. Image analysis in metamorph: Images were acquired using the IXM XLS automated microscope. One to four images per field were acquired. Exposure times were adjusted per sample to have the greatest dynamic range of fluorescent signal. For each screen on a cell/model/patient sample, the exposure time was kept constant. Cytochrome c quantification was performed using the multiwavelength cell scoring module in metamorph. Briefly, this module segments cells based on an intensity above local background of cytochrome c. This results in an approximate single cell segmentation (for example, as shown in fig S7) and a quantification of the area of cytochrome c staining. Cells were scored as positive or negative based on whether the area exceeded a user-specified cutoff. Cutoffs were identified by using known positive- and negative-control drugs. Typically, several cutoffs were selected, and the best cutoff was identified from the screen. Quantification of delta priming: The percentage of cytochrome c positive cells in a well was translated into delta priming measurements as described in the text. Briefly, the percentage of cytochrome c positive cells were calculated for all DMSO wells, and the difference between positive cells in DMSO and drug treated wells was calculated. To calculate delta priming, we use the percent positive cytochrome c cells in DMSO wells from each plate (and not across the whole experiment). This can mitigate plate effects in screening.

𝐷𝑒𝑙𝑡𝑎 𝑃𝑟𝑖𝑚𝑖𝑛𝑔 = % 𝑃𝑜𝑠𝑖𝑡𝑖𝑣𝑒 𝐶𝑦𝑡𝑜𝑐ℎ𝑟𝑜𝑚𝑒 𝑐 𝐶𝑒𝑙𝑙𝑠𝐷𝑀𝑆𝑂 𝑇𝑟𝑒𝑎𝑡𝑒𝑑 𝑊𝑒𝑙𝑙𝑠

− % 𝑃𝑜𝑠𝑖𝑡𝑖𝑣𝑒 𝐶𝑦𝑡𝑜𝑐ℎ𝑟𝑜𝑚𝑒 𝑐 𝐶𝑒𝑙𝑙𝑠𝐶𝑜𝑚𝑝𝑜𝑢𝑛𝑑 𝑇𝑟𝑒𝑎𝑡𝑒𝑑 𝑊𝑒𝑙𝑙𝑠

For each well, we identified the delta priming value. These raw values were matched to the well map using excel. Hits were identified as compounds that increase apoptotic priming at least 3 standard deviation above the background seen in DMSO treated wells. Comparisons of delta priming between different tumors is complicated by two factors: (i) assay noise in present in the DMSO treated wells, and (ii) the maximum level of cytochrome c observed. DMSO served as the negative control in our screen. Using typical screening approaches, hits were identified as compounds that have a delta priming signal at least 3 times greater than the standard deviation observed in the negative control. In different colon PDX models, we observed different hit cutoffs (fig. S11A). Furthermore, we observe that there is a different maximum delta priming that we can measure for each tumor. This is a result of the amount of cytochrome c loss present in DMSO-treated wells which itself is dependent on the viability of cells after 24 hours, and the baseline cytochrome c release at the screening peptide concentration. To enable a comparison of delta priming of different PDX models, we performed a linear normalization of delta priming so that 0 corresponds to 3 times the standard deviation of DMSO treated wells and 100 corresponds to the maximum delta priming value measured. All other compounds that do not increase delta priming above the noise cutoff are set to 0. The formula for normalized delta priming is described below:

𝒊𝒇 𝐷𝑒𝑙𝑡𝑎 𝑃𝑟𝑖𝑚𝑖𝑛𝑔 > 3 × 𝜎𝐷𝑀𝑆𝑂 𝑊𝑒𝑙𝑙𝑠 𝒕𝒉𝒆𝒏 𝑁𝑜𝑟𝑚𝑎𝑙𝑖𝑧𝑒𝑑 𝐷𝑒𝑙𝑡𝑎 𝑃𝑟𝑖𝑚𝑖𝑛𝑔 =

𝐷𝑒𝑙𝑡𝑎 𝑃𝑟𝑖𝑚𝑖𝑛𝑔 − 3 × 𝜎𝐷𝑀𝑆𝑂 𝑊𝑒𝑙𝑙

𝐷𝑒𝑙𝑡𝑎 𝑃𝑟𝑖𝑚𝑖𝑛𝑔 𝑀𝑎𝑥𝑖𝑚𝑢𝑚𝐶𝑜𝑚𝑝𝑜𝑢𝑛𝑑

− 3 × 𝜎𝐷𝑀𝑆𝑂 𝑊𝑒𝑙𝑙

𝒊𝒇 𝐷𝑒𝑙𝑡𝑎 𝑃𝑟𝑖𝑚𝑖𝑛𝑔 < 3 × 𝜎𝐷𝑀𝑆𝑂 𝑊𝑒𝑙𝑙𝑠 𝒕𝒉𝒆𝒏 𝑁𝑜𝑟𝑚𝑎𝑙𝑖𝑧𝑒𝑑 𝐷𝑒𝑙𝑡𝑎 𝑃𝑟𝑖𝑚𝑖𝑛𝑔

= 0

Fig. S1. Schematic of DBP. (A) BH3 profiling measures the proximity of a cell to the apoptotic threshold.

(B) DBP measures whether a brief ex vivo treatment with a chemical sensitizes cells for apoptosis. (C)

Workflow for DBP. Cells are incubated with the drugs for 6-24 hours, cells are then permeabilized to

expose mitochondria, and are subsequently treated with synthetic BH3 peptides modeled on the Bcl-2

family of proteins. At a single intermediate peptide concentration, if the drug treatment has no impact on

apoptotic sensitivity, the cells retain cytochrome c in their mitochondria. Conversely if a drug increases

apoptotic sensitivity of the cell, the peptide treatment results in loss of cytochrome c from mitochondria.

We typically measure MOMP by immunofluorescent staining of endogenous cytochrome c. (D) Previous

FACS based dynamic BH3 profiling workflow. Earlier technological iterations of dynamic BH3 profiling

involved treating cancer cells with drugs, trypsinizng and counting them, then putting cells into BH3

profiling buffers. This earlier iteration required more human handling and was not amenable to high-

throughput screening.

Fig. S2. Selection of informative peptide screening concentrations. The ideal peptide screening

concentration using for the BH3 profiling assay is selected such that if the cells are treated with DMSO (or

other chemical library vehicles), and subsequently BH3 profiled at the ideal concentration, then the

mitochondria in cells will retain cytochrome c. Conversely, if cells are treated with a chemical that

sensitizes cells for apoptosis prior to BH3 profiling, then the subsequent BH3 profile at the ideal peptide

concentration will result in loss of cytochrome c from mitochondria. The loss of cytochrome c from

mitochondria is represented by the delta priming metric. Identification of the ideal peptide concentration

is accomplished by treating a plate of cells with DMSO (referred to as the pre-profile plate) and treating

the remaining cells with the chemical library.

Fig. S3. Identification of a drug screening concentration. In working with tumor samples, we are

limited by the number of concentrations or drugs that are testable. We therefore selected one drug

concentration to work with in the course of these studies. We found that in performing HT-DBP on an ad

hoc library of 80 compounds that concentrations of 0.12 µM, 0.37µM and 1.1µM produced similar delta

priming results. In contrast, screening at a concentration of 3.3µM produced an increased number of hits

suggesting off target effects. All r values represent Pearson correlation coefficients. Data is mean of 2

independent experiments.

Fig. S4. HT-DBP enables identification of compounds that sensitize cancer cells for apoptosis. (A)

Images of cytochrome c loss in MDA-MB-231 breast cancer cells with increasing concentrations of the

synthetic Bim BH3 peptide. Cytochrome c immunofluorescence in green, Hoechst 33342 in blue. Scale

bar represents 100 microns. Images are representative of 2 independent experiments. (B) Quantification

of peptide induced cytochrome c loss in MDA-MB-231 cells. Data is representative of 2 independent

experiments. Error bars represent SD of technical replicates. Arrow indicates the BH3 peptide

concentration at which the BH3 profile is performed. (C) HT-DBP in MDA-MB-231 breast cancer cells

using 1902 compounds from the ICCB Selleck Bioactive Library. Compound treated wells shown in black.

DMSO treated wells shown in blue. Data represent mean of n=2 independent experiments. (D) Images

of annexin V staining used to measure cell death in MDA-MB-231 cells. Images are representative of 2

independent experiments. (E) Comparison of HT-DBP in MDA-MB-231 breast cancer cells at 24 hours

with frank cell death measured by annexin V staining at 72 hours (R2 = 0.50; p < 0.001; Pearson). Data

are mean of 2 independent experiments.

Fig. S5. Similarity between FACS and microscopy dynamic BH3 profiles. (A) FACS based dynamic

BH3 profiling in Panc8902 cells. Cells treated with 1µM of drug for 24 hours, trypsinized and

subsequently BH3 profiled. (B) Microscopy based dynamic BH3 profiles in Panc8902 cells. Cells treated

with 1µM of drugs for 24 hours and subsequently BH3 profiled. (C) Comparison of delta priming using the

FACS and microscopy-based methods show that there is a similar drug ranking when using FACS or

microscopy based dynamic BH3 profiles. (R2 = 0.82; p < 0.001; Pearson). Data is representative of 3

independent experiments.

Fig. S6. Images of stained MMTV-PyMT tumor cells. After BH3 profiling, MMTV-PyMT tumor cells were

stained with Hoechst 33342 (blue; DNA/nuclei), mouse FITC-conjugated CD326 (EpCAM) antibody

(green), and Alexa-647-conjugated cytochrome c antibody. A few cells are Epcam-negative, indicated by

white arrows. Scale bars, 100 µm.

Fig. S7. Example of cell masks. This is an example of the identification of nuclei and cytochrome c

staining of MMTv-PyMT tumor cells using the multi-wavelength cell scoring module in Metamorph. In the

merged image, blue represents Hoechst 33342 (nuclei), green represents EpCam, and red represents

cytochrome c. Cell boundaries were identified by cytochrome c signal.

Fig. S8. Screening data for freshly isolated MMTV-PyMT tumors. (A) Pre-screen Z’ scores of MMTV-PyMT tumors indicate negligible positional plate bias. (B) Z’ score from different plates for a screen of MMTV-PyMT tumors. DMSO is used as the negative control, 1 µM of AZD8055 is used as the positive control. Data is representative of 2 independent experiments. (C) Technical replicates of chemical

screens (R2 = 0.81; p < 0.001, by Pearson analysis from a single screen). (D) Biological replicates of

chemical screens (R2 = 0.81; p < 0.0001, by Pearson analysis).

Fig. S9. Counterscreens in healthy cells. (A) Bid peptide dose response curves in freshly isolated

mouse hepatocytes. Scale bars, 200 µm. Images are representative of n=2 screens. (B) Quantification of

the dose response curve in (A). Data is representative of n=2 screens. (C) HT-DBP on adult mouse

hepatocytes to identify drugs that increase apoptotic priming in hepatocytes. Data is mean of n=2

screens. (D) Bim peptide dose response curves in human foreskin fibroblasts. Scale bar, 100 µm. Images

are representative of n=2 screens. (E) Quantification of the dose response curve in (D). Data is

representative of n=2 screens. (F) HT-DBP on human foreskin fibroblasts shows several drugs that

increase apoptotic priming in fibroblasts. Data is mean of n=2 screens.

Fig. S10. Supplemental data for in vivo validation experiments for HT-DBP. (A) Images of Bim

peptide dose response curve in the BM-355 breast cancer PDX model. Scale bars, 200 µm. (B)

Quantification of the Bim peptide dose response curve in BM-355. For this dynamic BH3 profile, a

concentration of 0.205 µM Bim (indicated by the arrow) was chosen to perform the dynamic BH3 profile.

Images and quantification are representative of 2 independent experiments.

Fig. S11. Identification of compounds that sensitize the COCA9 colon cancer PDX for apoptosis.

(A) Example of staining by Hoechst 33342, Alexa488-conjugated antibody to EpCam, and Alexa647-

conjugated antibody to cytochrome c in COCA9 cells. Scale bars, 200 µm. (B) Images of cytochrome c

loss in COCA9 cells with increasing concentrations of the synthetic Bim BH3 peptide. Scale bars, 200 µm.

(C) Quantification of peptide-induced cytochrome c loss in COCA9 cells described in (B). Data represent

means ±SD. Arrow indicates the peptide screening concentration. (D) HT-DBP screen in COCA9 cells

using 1650 compounds from the Selleck Bioactive Library (table S1).

Fig. S12. Quantification of apoptotic chemical vulnerabilities in colon cancer PDX models. (A)

Demonstration that different HT-DBP screens in colon cancer PDX models have variable standard

deviation of delta priming in DMSO treated wells. (B) Comparison of priming across cell lines at 3 times

the standard deviation of DMSO-treated wells, as is often used as a cutoff for hits in chemical screens.

Variability of the maximum delta priming, and the cutoff priming value for a chemical hit makes tumor to

tumor comparisons difficult and requires normalization. COCA, colorectal adenocarcinoma lines; HFF,

human foreskin fibroblasts; mHepato, mouse hepatocytes.

Fig. S13. Heat map of apoptotic chemical vulnerabilities in colon cancer PDX models. Heat map of

normalized delta priming of all compounds for colon PDX models and healthy cells. Compounds are

organized from top to bottom based on average normalized delta priming of the colon PDX models.

Compounds and results are detailed in data file S10.

Fig. S14. Comparison of apoptotic chemical vulnerabilities in colon cancer PDX models.

Comparison of delta priming for specific classes of inhibitors across different PDX models (N indicated by

number of dots in the plots). Asterisks indicate adjusted P value <0.05 by Dunn’s multiple comparison test

(COCA39: adjusted P=0.002 pan-CDK, P>0.99 pan-HDAC; COCA74M: P>0.99 pan-CDK, P=0.0017 pan-

HDAC;).

Fig. S15. Supplemental data for comparison of chemical vulnerabilities of freshly isolated and

cultured cancer cells. (A) Scheme for generating MMTV-PyMT cancer cell lines, and for the

experimental comparison of chemical vulnerabilities using HT-DBP. (B) Delta priming and cell counts

from a HT-DBP screen on a cell line derived from the MMTV-PyMT tumor model. Data represent the

mean of n=2 independent experiments. (C) Evaluation of the area under curve of drug dose responses

for the MMTV-PyMT derived cell line and cells freshly isolated from MMTV-PyMT tumors. Data represent

the mean ±SD of n=2 independent experiments.

Fig. S16. Supplemental data for comparison of chemical vulnerabilities of freshly isolated and

cultured cancer cells. Evaluation of HT-DBP on COCA74P and COCA61 at 3 drug concentrations (100

nM, 250 nM, and 1000 nM) applied on day 1 and day 8 after tumor extraction. Data are means of at least

two replicates. Data behind the heat map are in data file S15.

Table S1. List of compounds used in the chemical screen. The table, provided as an xls file online, lists

the compounds tested in the screens.

Data file S1. Cell count and delta priming of MDA-MB-231 line. Total cell count and delta priming of

MDA-MB-231 cells at 24 hours. Data are mean of independent experiments and correspond to fig. S4C;

further details and analysis therein.

Data file S2. Apoptosis and delta priming of MDA-MB-231 line. Annexin V-positive cells at 72 hours

and delta priming at 24 hours of MDA-MB-231 cells. Data are mean of independent experiments and

correspond to fig. S4E; further details and analysis therein.

Data file S3. Cell count and delta priming of MMTV-PyMT tumor. Total cell count and delta priming of

freshly isolated MMTV-PyMT tumor cells at 24 hours. Data are mean of independent experiments and

correspond to Fig. 1D; further details and analysis therein.

Data file S4. Technical replicate of MMTV-PyMT delta priming. A representative technical replicate of

delta priming of freshly isolated MMTV-PyMT tumor cells at 24 hours. Data is representative of

independent experiments and correspond to fig. S8C; further details and analysis therein.

Data file S5. Biological replicate of MMTV-PyMT delta priming. A representative biological replicate of

delta priming of freshly isolated MMTV-PyMT tumor cells at 24 hours. Data corresponds to fig. S8D;

assay and analytical details therein.

Data file S6. Cell count and delta priming of primary mouse hepatocytes. Total cell count and delta

priming of freshly isolated mouse hepatocytes at 24 hours. Data are mean of independent experiments

and correspond to fig. S9C; further details and analysis therein.

Data file S7. Cell count and delta priming of human foreskin fibroblasts. Total cell count and delta

priming of human foreskin fibroblast cell line at 24 hours. Data are mean of independent experiments

and correspond to fig. S9E; further details and analysis therein.

Data file S8. Delta priming in adult mouse hepatocytes and MMTV-PyMT tumors. Delta priming of

adult mouse hepatocytes and freshly isolated MMTV-PyMT mouse tumors at 24 hours. Data are mean

of independent experiments and corresponds to Fig. 1F; further details and analysis therein.

Data file S9. Delta priming of freshly isolated MMTV-PyMT tumors by nominal drug target. Delta

priming of freshly isolated MMTV-PyMT mouse tumors at 24 hours organized according to nominal drug

target. Note that compounds that have multiple targets are listed multiple times. Data are mean of

independent experiments and correspond to Fig. 1G; further details and analysis therein.

Data file S10. Normalized delta priming of colorectal PDX models. Data correspond to Fig. 3, A and C,

and fig. S13; assay and analytical details therein.

Data file S11. Normalized delta priming of colorectal PDX models by nominal target. Normalized delta

priming of colorectal PDX models at 24 hours organized according to nominal target. Data corresponds

to Fig. 3C; assay and analytical details therein.

Data file S12. Delta priming of MMTV-PyMT–derived cell line. Delta priming at 24 hours of a cell line

derived from MMTV-PyMT tumors. Data represent mean of independent experiments and corresponds

to fig. S15B; further assays and analytical details therein.

Data file S13. Correlation between MMTV-PyMT tumor and derived cell line. Correlation of delta

priming at 24 hours in freshly isolated MMTV-PyMT tumor cells and in a derived cell line. Data are mean

of independent experiments and correspond to Fig. 4A; further details and analysis therein.

Data file S14. Dose response by delta priming in MMTV-PyMT tumor and derived cell line. Delta

priming of listed drugs by dose in freshly isolated MMTV-PyMT tumor cells and a derived cell line. Data

represents the area under curve of delta priming. Data calculated as mean of independent experiments

and corresponds to fig. S15C; further details and analysis therein.

Data file S15. Delta priming in colorectal PDX models. Delta priming of several drugs at 3

concentrations for colorectal PDX models at 24 hours after tumor extraction and 8 days after tumor

extraction. Data corresponds to Fig. 4, F and G, and fig. S16; assay and analytical details therein.

Data file S16. Delta priming in primary human CRC cells ex vivo. Delta priming of several drugs at

multiple doses for primary human colorectal samples at 24 hours after tumor extraction. Data

corresponds to Fig. 5C; assay and analytical details therein.