Dataset DOI: 10.5061/dryad.h70rxwdzs

Description of the data and file structure

Data from peer-reviewed article:

Title: Control of the Integrated Stress Response Limits Glioblastoma Invasion

Journal: Cell Biochemistry and Function (manuscript ID 1944219)

Authors: Lisa K. Månsson, Ethan Dickson, Lun Hao, Angela A. Pitenis, Maxwell Z. Wilson

Corresponding authors: Angela A. Pitenis, apitenis@ucsb.edu and Maxwell Z. Wilson, mzw@ucsb.edu

Files and variables

File: 2026-04-11-Mansson-DataDryad.zip

File List

1. fig1d_i.png
2. fig1d_ii.png
3. fig1d_iii.png
4. fig1d_iv.png
5. fig1e_repeat1.csv
6. fig1e_repeat2.csv
7. fig1e_repeat3.csv
8. fig1f_repeat1.csv
9. fig1g_repeat1.csv
10. fig2_countfiles_dark_1.txt
11. fig2_countfiles_dark_2.txt
12. fig2_countfiles_light_1.txt
13. fig2_countfiles_light_2.txt
14. fig2_countfiles_light_3.txt
15. fig2_Rscript.txt
16. fig2_LFCresults_gene_list.csv
17. fig3b.csv
18. fig3c_ECMstain_1.csv
19. fig3c_ECMstain_2.csv
20. fig3c_ECMstain_3.csv
21. fig3ef.csv
22. fig4b_i_WT_light_0h.tif
23. fig4b_i_WT_light_6h.tif
24. fig4b_i_WT_light_12h.tif
25. fig4b_i_WT_light_18h.tif
26. fig4b_i_WT_light_23h.tif
27. fig4b_ii_opto_dark_0h.tif
28. fig4b_ii_opto_dark_6h.tif
29. fig4b_ii_opto_dark_12h.tif
30. fig4b_ii_opto_dark_18h.tif
31. fig4b_ii_opto_dark_23h.tif
32. fig4b_iii_opto_light_0h.tif
33. fig4b_iii_opto_light_6h.tif
34. fig4b_iii_opto_light_12h.tif
35. fig4b_iii_opto_light_18h.tif
36. fig4b_iii_opto_light_23h.tif
37. fig4cde.csv
38. fig5a_i_wt_light.png
39. fig5a_i_opto_dark.png
40. fig5a_i_opto_light.png
41. fig5a_ii_wt_light.png
42. fig5a_ii_opto_dark.png
43. fig5a_ii_opto_light.png
44. fig5a_iii_wt_light.png
45. fig5a_iii_opto_dark.png
46. fig5a_iii_opto_light.png
47. fig5bcd.csv
48. fig6a_LEFT_22h.tif
49. fig6a_RIGHT_22h.tif
50. fig6b.csv
51. fig7a_i_0h.png
52. fig7a_i_24h.png
53. fig7a_i_72h.png
54. fig7a_ii_0h.png
55. fig7a_ii_24h.png
56. fig7a_ii_72h.png
57. fig7a_iii_0h.png
58. fig7a_iii_24h.png
59. fig7a_iii_72h.png
60. fig7bc.csv
61. SI_fig1_G3BP.png
62. SI_fig1_peIF2a.png
63. SI_fig1_ATF4.png
64. SI_fig1_CHOP.png
65. SI_fig2_G3BP.png
66. SI_fig2_peIF2a.png
67. SI_fig2_ATF4.png
68. SI_fig2_CHOP.png
69. SI_fig5a_ECMstain_2.csv
70. SI_fig5b.csv
71. SI_fig6a_DAPI_opto_dark.png
72. SI_fig6a_DAPI_opto_light.png
73. SI_fig6a_DAPI_opto_light_ISRIB.png
74. SI_fig6a_DAPI_wt_light.png
75. SI_fig6a_SiriusRed_opto_dark.png
76. SI_fig6a_SiriusRed_opto_light.png
77. SI_fig6a_SiriusRed_opto_light_ISRIB.png
78. SI_fig6a_SiriusRed_wt_light.png
79. SI_fig6a_FastGreen_opto_dark.png
80. SI_fig6a_FastGreen_opto_light.png
81. SI_fig6a_FastGreen_opto_light_ISRIB.png
82. SI_fig6a_FastGreen_wt_light.png
83. SI_fig6b_0h_opto_dark.png
84. SI_fig6b_0h_opto_light.png
85. SI_fig6b_0h_opto_light_ISRIB.png
86. SI_fig6b_0h_wt_light.png
87. SI_fig6b_72h_opto_dark.png
88. SI_fig6b_72h_opto_light.png
89. SI_fig6b_72h_opto_light_ISRIB.png
90. SI_fig6b_72h_wt_light.png
91. SI_fig7_SF295_wt_light_0h.png
92. SI_fig7_SF295_opto_dark_0h.png
93. SI_fig7_SF295_opto_light_0h.png
94. SI_fig7_U251_wt_light_0h.png
95. SI_fig7_U251_opto_dark_0h.png
96. SI_fig7_U251_opto_light_0h.png
97. SI_fig7_H4_wt_light_0h.png
98. SI_fig7_H4_opto_dark_0h.png
99. SI_fig7_H4_opto_light_0h.png

Data description for non-image files

FIGURE 1: Opto-PKR allows for high spatiotemporal control with light-triggered ISR activation to examine its role in GBM cancer progression.

Description:

(a) GBM is a severe type of brain cancer with aggressive spread, finger‐like protrusions, and single cancer cells migrating long distances from their origin. (b) ISR is a complex signaling network that reprograms translational output in the cell via phosphorylation of eIF2α by the PKR kinase, followed by activation of transcription factors ATF4 and CHOP, through which the cell adapts and responds to the stress applied. ISR activation has been suggested to promote but also to suppress GBM progression in different studies. (c) Using an optogenetic tool for ISR activation developed by Batjargal et al. (2023), we achieve reduced molecular complexity and high spatiotemporal control of PKR‐kinase activation, inducing artificial ISR activation in cells. (d) Immunofluorescence images of downstream ISR‐markers in H4 neuroglioma cells after different amounts of ISR activation (time under light). (e–g) Quantification plots based on immunofluorescence images of cytoplasmic α‐p‐eIF2α, nuclear α‐ATF4, and nuclear α‐CHOP, for three different cell lines: H4 (e), SF‐295 (f), and U251 (g). All fluorescence values are normalized to the dark condition average for each protein. More than 100 cells were quantified for each time point for each cell line to obtain means (solid line) and standard deviation (shaded regions). Representative immunofluorescence images for SF‐295 and U251 are shown in Supporting Information Figures 1 and 2.

E) fig1e_repeat1.csv: CellPose data from IF on H4 cells with different amount of time under light-stimulating opto-PKR. Detected nuclear and cytoplasmic mean intensities per cell from the different antibodies and stains used (ch3: nuclear CHOP, ch4: cytoplasmic p-eIF2a or nuclear ATF4 - since both are on the same channels, these two stains are quantified in different images)

Column descriptions:

• frame_cyto - image frame number in which the cell from which cytoplasmic p-eIF2a was quantified (each row represents a cell) was detected.
• cyto_peIF2a_intensity - mean cytoplasmic intensity from p-eIF2a IF (a.u.)
• frame_nuclear - image frame number in which the cell from which nuclear CHOP as well as nuclear ATF4 was quantified (each row represents a cell) was detected.
• nuclear_CHOP_mean - mean nuclear intensity from CHOP IF (a.u.)
• nuclear_ATF4_mean - mean nuclear intensity from ATF4 IF (a.u.)

• row 1: Data header (frame_cyto, cyto_peIF2a_intensity, frame_nuclear, nuclear_CHOP_mean,nuclear_ATF4_mean). For repeat 1 (fig1e_repeat1.csv), there are 3 frames per time condition in this file, e.g., frames 0-2 are opto-dark, 3-5 are 5 min light exposure, etc.
• row 2 to end: Data - each row represents one cell. Note that the cells stained with G3BP and p-eIF2a are not the same cells as those stained with CHOP and ATF4, why there is a separate column with frame numbers (frame_nuclear) for nuclear CHOP and ATF4.

fig1e_repeat2.csv: CellPose data from IF on H4 cells with different amount of time under light-stimulating opto-PKR. Detected nuclear and cytoplasmic mean intensities per cell from the different antibodies and stains used (ch3: nuclear CHOP, ch4: cytoplasmic p-eIF2a or nuclear ATF4)

Column descriptions:

• frame_cyto - image frame number in which the cell from which cytoplasmic p-eIF2a was quantified (each row represents a cell) was detected.
• cyto_peIF2a_intensity - mean cytoplasmic intensity from p-eIF2a IF (a.u.)
• frame_nuclear - image frame number in which the cell from which nuclear CHOP as well as nuclear ATF4 was quantified (each row represents a cell) was detected.
• nuclear_CHOP_mean - mean nuclear intensity from CHOP IF (a.u.)
• nuclear_ATF4_mean - mean nuclear intensity from ATF4 IF (a.u.)

• row 1: Data header (frame_cyto, cyto_peIF2a_intensity, frame_nuclear, nuclear_CHOP_mean,nuclear_ATF4_mean). For repeat 2 (fig1e_repeat2.csv) there are 5 frames per time condition, so frame 0-4 is for dark, 5-8 is for 5 min light, etc.
• row 2 to end: Data - each row represents one cell. Note that the cells stained with G3BP and p-eIF2a are not the same cells as those stained with CHOP and ATF4, why there is a separate column with frame numbers (frame_nuclear) for nuclear CHOP and ATF4.

fig1e_repeat3.csv: CellPose data from IF on H4 cells with different amount of time under light-stimulating opto-PKR. Detected nuclear and cytoplasmic mean intensities per cell from the different antibodies and stains used (ch3: nuclear CHOP, ch4: cytoplasmic p-eIF2a or nuclear ATF4)

Column descriptions:

• frame_cyto - image frame number in which the cell from which cytoplasmic p-eIF2a was quantified (each row represents a cell) was detected.
• cyto_peIF2a_intensity - mean cytoplasmic intensity from p-eIF2a IF (a.u.)
• frame_nuclear - image frame number in which the cell from which nuclear CHOP as well as nuclear ATF4 was quantified (each row represents a cell) was detected.
• nuclear_CHOP_mean - mean nuclear intensity from CHOP IF (a.u.)
• nuclear_ATF4_mean - mean nuclear intensity from ATF4 IF (a.u.)

• row 1: Data header (frame_cyto, cyto_peIF2a_intensity, frame_nuclear, nuclear_CHOP_mean,nuclear_ATF4_mean). For repeat 3 (fig1e_repeat3.csv) there are 4 frames per time condition, so frame 0-3 is for dark, frame 4-7 is for 5 min light etc.
• row 2 to end: Data - each row represents one cell. Note that the cells stained with G3BP and p-eIF2a are not the same cells as those stained with CHOP and ATF4, why there is a separate column with frame numbers (frame_nuclear) for nuclear CHOP and ATF4.

F) fig1f_repeat1.csv: CellPose data from IF on U251 (NCI60 5.3a) cells with different amount of time under light-stimulating opto-PKR. Detected nuclear and cytoplasmic mean intensities per cell from the different antibodies and stains used (ch3: nuclear CHOP, ch4: cytoplasmic p-eIF2a or nuclear ATF4)

Column descriptions:

• frame_cyto - image frame number in which the cell from which cytoplasmic p-eIF2a was quantified (each row represents a cell) was detected.
• cyto_peIF2a_intensity - mean cytoplasmic intensity from p-eIF2a IF (a.u.)
• frame_nuclear - image frame number in which the cell from which nuclear CHOP as well as nuclear ATF4 was quantified (each row represents a cell) was detected.
• nuclear_CHOP_mean - mean nuclear intensity from CHOP IF (a.u.)
• nuclear_ATF4_mean - mean nuclear intensity from ATF4 IF (a.u.)

• row 1: Data header (frame_cyto, cyto_peIF2a_intensity, frame_nuclear, nuclear_CHOP_mean,nuclear_ATF4_mean). For repeat 1 (fig1e_rep1.csv) there are 4 frames per time condition in this file, e.g., frames 0-3 are opto-dark, 4-7 are 5 min light exposure, etc.
• row 2 to end: Data - each row represents one cell. Note that the cells stained with G3BP and p-eIF2a are not the same cells as those stained with CHOP and ATF4, why there is a separate column with frame numbers (frame_nuclear) for nuclear CHOP and ATF4.

G) fig1g_repeat1.csv: CellPose data from IF on SF-295 (NCI60 5.3b) cells with different amount of time under light stimulating opto-PKR. Detected nuclear and cytoplasmic mean intensities per cell from the different antibodies and stains used (ch3: nuclear CHOP, ch4: cytoplasmic p-eIF2a or nuclear ATF4)

Column descriptions:

• frame_cyto - image frame number in which the cell from which cytoplasmic p-eIF2a was quantified (each row represents a cell) was detected.
• cyto_peIF2a_intensity - mean cytoplasmic intensity from p-eIF2a IF (a.u.)
• frame_nuclear - image frame number in which the cell from which nuclear CHOP as well as nuclear ATF4 was quantified (each row represents a cell) was detected.
• nuclear_CHOP_mean - mean nuclear intensity from CHOP IF (a.u.)
• nuclear_ATF4_mean - mean nuclear intensity from ATF4 IF (a.u.)

• row 1: Data header (frame_cyto, cyto_peIF2a_intensity, frame_nuclear, nuclear_CHOP_mean,nuclear_ATF4_mean). For repeat1 (fig1e_rep1.csv) there are 4 frames per time condition in this file, e.g. frame 0-3 are opto-dark, 4-7 are 5 min light exposure, etc.
• row 2 to end: Data - each row represents one cell. Note that the cells stained with G3BP and p-eIF2a are not the same cells as those stained with CHOP and ATF4, why there is a separate column with frame numbers (frame_nuclear) for nuclear CHOP and ATF4.

FIGURE 2: Differential gene expression analysis after 24 h stimulation of opto-PKR H4 cells in light compared to control cells 24 h in dark.

Description:

(a) Differential gene expression analysis was performed after 24 h stimulation of opto‐PKR H4 cells in light (N = 3) compared to control cells 24 h in dark (N = 2). (b) Volcano plot presents differentially expressed genes from RNA‐sequencing data: a total of 1905 genes were differentially expressed out of the 11,982 genes with nonzero total read count, for an adjusted p‐value < 0.05. Nonsignificant differentially expressed genes are colored in light gray, all differentially expressed genes with an adjusted p‐value larger than 0.05 are colored dark gray, and genes with p‐value criteria as well as a log2fold change larger than 1.8, are colored black. (c) Heatmap of the 25 most up‐ and down‐regulated genes with significant enrichment analysis categories annotated on the right‐hand side. ISR activation for 24 h shows a trend for enrichment of genes related to ISR activation and cell signaling. Significant GO‐terms for enrichment on downregulated genes after ISR activation for 24 h are related to the extracellular matrix as well as glial cell migration.

B, C) fig2_countfiles_dark_1.txt, fig2_countfiles_dark_2.txt, fig2_countfiles_light_1.txt, fig2_countfiles_light_2.txt, fig2_countfiles_light_3.txt: Bulk RNA-sequencing data (raw counts) for opto-PKR H4 cells in 24h light (N=3 wells) and opto-PKR H4 cells in dark (N=2 wells).

• Col 1: GENEID: Gene name
• Col 10: 10: Count for the specific gene (row)

The rest of the columns are auto-generated 0-filled columns without data.

fig2_Rscript.txt: R-script with DESeq2 differential gene expression analysis for data processing according to the methods.

fig2_LFCresults_gene_list.csv: resulting gene list from running shrunken DEseq2 in R (file fig2_Rscript.txt).

• col 1: gene_name: gene name
• col 2: baseMean: average of the normalized count values for a gene, calculated across all samples in the dataset
• col 3: lfcSE: log2 fold change standard error
• col 4: stat: the Wald statistic (or likelihood ratio test statistic) for each gene, calculated as the shrunken log2FoldChange divided by its standard error (lfcSE). It is a z-statistic used to calculate the p-value by comparing the result to a standard normal distribution.
• col 5: pvalue: p-value
• col 6: padj: adjusted p-value

FIGURE 3: ISR activation over 72 h shows reduced collagen content in the extracellular environment and inhibited cellular spread.

Description:

In all subfigures, opto‐PKR H4 cells with ISR activation (“opto light”) are compared to wild‐type cells in light (“WT light”), opto‐PKR cells without ISR activation (“opto dark”), and opto‐PKR cells in light with inhibited ISR activation (“opto light +ISRIB”). (a) Procedure for 72 h ECM stain assay. (b) Quantification of cell viability after 72 h of ISR activation based on DAPI stain area coverage. Each data point represents quantification of a whole well in a 24‐well plate. (c) Quantification of fibrillar collagen to non‐collagen ECM ratio after 72 h of stimulation. Each data point represents quantification of a whole well in a 24‐well plate. (d) Procedure for 72 h scratch assay. (e) ISR activation inhibits the spread of cells, quantified from scratch closure distance in the scratch assay over 72 h (representative images shown in Supporting Information Figure 6). Each data point represents the average scratch closure in one well of a 12‐well plate. (f) Time‐resolved quantification of cellular spread in scratch assay from 0 h, to 24 h, to 48 h, up to 72 h of stimulation compared to controls. For all the subfigures, boxplots and barplots indicate mean values with error bars showing standard deviations (*p < 0.05, **p < 0.01, ***p < 0.001).

B) fig3b.csv: DAPI area coverage data at 72 hours - area of nuclei detected by StarDist (Fiji, ImageJ)

• col 1: sample: H4 cells of type opto-PKR or WT, in light or dark and with ISRIB.
• col 2: area_fraction_DAPI: area fraction DAPI stained nuclei take up in the well imaged (unitless).

C) fig3c_ECMstain_X.csv: ECM stain kit intensity readouts from plate reader in Cytation 5 in 540 and 605 nm for repeat X.

• col 1: sample: indicates what sample is represented on each row
• col 2,3: 72h: Samples that were in light for the whole 72h of the experiment.
• col 4,5: 0h: Samples for dark control (0 hours in light)
• col 6: readout_channel: wavelength of read in plate reader (nm)

E, F) fig3ef.csv: Scratch assay quantification. csv-file with all measured scratch reductions in µm after 0h (0 reduction for all), 24 h, 48 h, and 72 h for all samples, and all repeats.

• row 1: sample: opto-PKR cells in light or dark and with ISRIB, WT cells
• row 2-5: time stamp (0-72 h)

FIGURE 4: Light-controlled ISR activation limits GBM spread in 3D.

Description:

(a) Spheroid formation protocol for 3D culture in collagen I matrix. (b) Maximum intensity projection images of confocal z-stacks of GBM spheroids. Time-lapse imaging over one day of ISR activation in opto-PKR cells (opto light) compared to controls; opto-PKR cells in dark (opto dark), and WT cells in light (WT light). (c–e) Quantifications of spheroid containment by area change, perimeter change, and circularity. Circles show data points, and shaded regions are standard deviations based on n = 3 technical replicates. Images and corresponding quantifications in this figure are representative data from experiments on the cell line SF-295 dyed with CellTracker Red.

C) fig4cde.csv: quantification data for spheroid perimeter and area across different samples and conditions in a time lapse of 22 hours.

• row 1: Repeat label (replicate 1,2,3)
• row 2: Labels for all data columns (Condition - WT/opto-PKR cells - light/dark, Time point (h), Area (µm2), Perimeter (µm))
• row 3-end: data

FIGURE 5: ISR activation limits GBM spread in 3D across different cell lines (SF-295 GBM, U251 GBM, H4 neuroglioma).

Description:

ISR activation limits GBM spread in 3D across different cell lines (SF-295 GBM, U251 GBM, H4 neuroglioma). Spheroids in a collagen matrix were imaged before and after stimulation with light or kept in the dark for 24 h. (a) Maximum intensity projections from confocal imaging of spheroid z-stacks after 24 h in collagen matrix. Controls (WT light, opto dark) are compared to ISR-activated spheroids (opto light). Figure (b–d) shows quantifications of spheroid containment after 24 h from maximum intensity projections of z-stacks from confocal imaging. Boxplots show mean values, with error bars indicating standard deviations. Each data point represents one spheroid. All cell lines examined show significant containment under ISR activation compared to controls (*p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001).

B, C, D) fig5bcd.csv: Spheroid quantification data.

• col 1: Cell_line: cell line (H4, SF-295, or U251)
• col 2: Condition: opto-PKR/WT cells in light/dark
• col 3: Area_0h: Area of spheroid at 0 hours (µm2)
• col 4: Perimeter_0h: Perimeter of spheroid at 0 hours (µm)
• col 5: Area_24h: Area of spheroid at 24 hours (µm2)
• col 6: Perimeter_24h: Perimeter of spheroid at 24 hours (µm)
• col 7: Area_72h: Area of spheroid at 72 hours (48 h recovery data) (µm2)
• col 8: Perimeter_72h: Perimeter of spheroid at 72 hours (48 h recovery data) (µm)

FIGURE 6: Photopatterned GBM containment in SF-295 spheroid obtained with precise spatiotemporal control of ISR activation.

Description:

(a) Using digital micromirror devices on a confocal microscope, only a rectangular region of interest was illuminated to spatially constrain the induction of ISR activation to one half of a spheroid. (b) Quantification of spheroid containment by the area change for each of the two half-spheres pictured in (a) (light vs. dark) over 22 h. Circles show data points, and shaded regions are standard deviations based on n = 3 technical replicates. The spheroid was dyed with CellTracker Red.

C) fig6b.csv: quantification data for spheroid area change across different samples and conditions in time lapse of 22 hours.

• row 1: Repeat label (replicate 1,2,3)
• row 2: Labels for all data columns (Condition - opto-PKR cells in light (right)/dark (left), Time point (h), Area (µm2), Perimeter (µm))
• row 3-end: data

FIGURE 7: Spheroids recover spread after ISR activation across multiple cell lines.

Description:

To study recovery after ISR activation, opto-PKR spheroids were illuminated for 24 h, imaged, and thereafter kept in the dark for 48 h and imaged again. (a) DIC images of opto-PKR spheroids at timepoints 0 h, 24 h, and 72 h. (b) and (c) Quantifications of spheroid spread recovery across all cell lines (SF-295, U251, H4) WT cells compared to opto-PKR cells.

B,C) fig7bc.csv: Spheroid quantification data.

• col 1: Cell_line: cell line (H4, SF-295, or U251)
• col 2: Condition: opto-PKR/WT cells in light/dark
• col 3: Area_0h: Area of spheroid at 0 hours (µm2)
• col 4: Perimeter_0h: Perimeter of spheroid at 0 hours (µm)
• col 5: Area_24h: Area of spheroid at 24 hours (µm2)
• col 6: Perimeter_24h: Perimeter of spheroid at 24 hours (µm)
• col 7: Area_72h: Area of spheroid at 72 hours (48 h recovery data) (µm2)
• col 8: Perimeter_72h: Perimeter of spheroid at 72 hours (48 h recovery data) (µm)

SUPPLEMENTARY INFORMATION

SI FIGURE 1:

Description:

Immunofluorescence imaging of fixed and stained SF-295 opto-PKR GBM cells in 24h light (pulsed 5s ON, 15s OFF) compared to time zero for proteins involved in ISR activation: phosphorylation of eIF2a, G3BP (in stress granules), ATF4, and CHOP. Quantification of immunofluorescence is shown in Fig. 1 in the main manuscript.

SI FIGURE 2:

Description:

Immunofluorescence imaging of fixed and stained U251 opto-PKR GBM cells in 24h light (pulsed 5s ON, 15s OFF) compared to time zero for proteins involved in ISR activation: phosphorylation of eIF2a, G3BP (in stress granules), ATF4, and CHOP. Quantification of immunofluorescence is shown in Fig. 1 in the main manuscript.

SI FIGURE 3:

Description:

MA plots for all genes (dark blue: differentially expressed genes, adjusted p-value < 0.05; gray: non-significant genes) from RNA-sequencing of opto-PKR H4 cells in 24h light compared to 24h dark-samples before (a) versus after (b) shrinkage using LFCshrink. Each dot in the MA plot represents one gene. The shrinkage algorithm performs a correction for high log fold change values on low count samples to avoid false positives, as can be seen comparing (a) and (b) above, where genes with low normalized count mean values have distinctly lowered log fold change values in (b).

Plots in Fig. 3 a) and b) as well as Fig.4 were generated from raw data and code under FIGURE 2 for main manuscript above.

SI FIGURE 4:

Description:

Pearson correlation heatmap between RNA-sequencing samples.

Plot in Fig.4 was generated from raw data and code under FIGURE 2 for main manuscript above.

SI FIGURE 5: ECM stain assay and DAPI coverage at earlier time points.

Description:

a) ECM-stain assay (see methods section) to examine collagen fraction for different amounts of time under stress, performed on H4 cells with 24, 48 and 72 hours under light (N=1 replicate per time point and condition). The total time for the experiment for all cells was 72 h, where the cells exposed for 24 h first grew for 48 h in the dark. Similarly, the cells exposed for 48 h had been growing in the dark for 24 h before. b) DAPI coverage in 96-well for opto-PKR H4 cells, seeded at 10 000 cells per well, cultured in the dark vs light for 24 hrs. Bar heights show the average of two replicates. The “opto”-label in both figures indicate cells with opto-PKR.

A) SI_fig5a_ECMstain_2.csv: ECM stain kit readouts from plate reader in Cytation 5 in 540 and 605 nm.

• col 1: sample: indicates what sample is represented on each row
• col 2,3: 72h: Samples that were in light for the whole 72h of the experiment.
• col 4: 48h: Samples that were in light for the last 48h out of the 72h of the experiment.
• col 5: 24h: Samples that were in light for the last 24h out of the 72h of the experiment.
• col 6,7: 0h: Samples for dark control
• col 8: readout_channel: wavelength of read in plate reader (nm)

B) SI_fig5b.csv: DAPI area coverage data at 24 hours.

• col 1: sample: H4 cells of type opto-PKR, in light or dark.
• col 2: area_fraction_DAPI: area fraction DAPI stained nuclei take up in the well imaged (unitless).

SI FIGURE 6:

Description:

a) Representative images of whole wells with H4 cells with nuclear stain (DAPI), as well as ECM stain (Sirius Red - collagenous protein, Fast Green - non-collagenous protein). b) Representative images of scratch assay for quantification plots in main Fig. 3.

SI FIGURE 7:

Description:

Representative images for each cell line (SF-295, U251, H4) showing spheroid containment after one day (24h) of ISR activation compared to controls. These images are the same as in Fig. 5 in the main manuscript but include images for each spheroid at time point 0 h.