Figure 2. Two branches of the TGFβ pathway display different signaling dynamics.

(A) BMP and ACTIVIN signaling represent the two branches of the TGFβ pathway. BMP signals through R-SMAD1 and ACTIVIN signals through R-SMAD2. R-SMADs form a complex with co-SMAD4 and the complex regulates target gene expression. (B–C) RUES2-RFP-SMAD1 grown on micropatterned colonies in E7 and stimulated with BMP4 (50 ng/mL). The response of cells at the colony edge (B) and near the colony center (C) as a function of time following BMP4 presentation. The intensity range was adjusted to the same minimum and maximum values in all images in both (B) and (C). Scale bars, 25 µm. (D) Average RFP-SMAD1 nuclear signal as a function of radial position within the colony at different time points following BMP4 treatment. The single-cell nuclear RFP-SMAD1 intensity was quantified and normalized to the single-cell H2B-mCitrine signal. Error bars represent the standard deviation over n = 5 colonies from one experiment. (E–F) RUES2-mCit-SMAD2 grown on micropatterned colonies in E7 and stimulated with ACTIVIN (10 ng/mL). The response of cells at the colony edge (E) and near the colony center (F) as a function of time following ACTIVIN presentation. The intensity range was adjusted to the same minimum and maximum values in all images in both (E) and (F). Scale bars represent 25 µm. (G) Average mCitrine-SMAD2 nuclear-to-cytoplasmic signal as a function of radial position within the colony at different time points following ACTIVIN treatment. The single-cell nuclear mCitrine intensity was quantified and normalized to the single-cell cytoplasmic mCitrine signal. Error bars represent the standard deviation over n = 5 colonies from one experiment.

Figure 2—figure supplement 1. R-SMAD reporter line generation.

(A) Schematic of the modification of the endogenous SMAD1 locus via CRISPR/Cas9 genome engineering. A blasticidin resistance gene (BsdR) is separated from the tagRFP-SMAD1 fusion by a T2A self-cleaving peptide. The SMAD1 coding sequence (e1) and the upstream untranslated regions are not affected by the modification. (B) Sequence analysis using loci specific primers (magenta arrows in A) demonstrated that both of the SMAD1 alleles had been modified. The trace shows the modified allele sequence. (C) RUES2-RFP-SMAD1 cells maintain expression of pluripotency markers under standard growth conditions. Separate samples were prepared for each marker and immunofluorescence staining was carried out using Alexa Fluor 647-conjgated secondary antibodies to avoid signal from the reporter constructs: NANOG, OCT4, and SOX2 (green) and DAPI (red). Scale bar, 50 µM. (D) RUES2-RFP-SMAD1 has a normal XX karyotype. (E) Schematic of the modification of the endogenous SMAD2 locus via CRISPR/Cas9 genome engineering. A puromycin resistance gene (PuroR) is separated from the mCitrine-SMAD2 fusion by a T2A self-cleaving peptide. The SMAD2 coding sequence (e1) and the upstream untranslated regions are not affected by the modification. (F) Sequence analysis using loci specific primers (magenta arrows in E) demonstrated that only one of the SMAD2 alleles had been targeted. The trace shows the modified allele sequence. (G) RUES2-mCit-SMAD2 cells maintain expression of pluripotency markers under standard growth conditions. Separate samples were prepared for each marker and immunofluorescence staining was carried out using Alexa Fluor 647-conjgated secondary antibodies to avoid signal from the reporter constructs: NANOG, OCT4, and SOX2 (green) and DAPI (red). Scale bar, 50 µM. (H) RUES2-mCit-SMAD2 has a normal XX karyotype.

Figure 2—figure supplement 2. Quantification of R-SMAD dynamics at the single-cell level in micropatterned colonies.

(A) Schematic of the image analysis pipeline for micropatterned culture experiments. The stitched and background-corrected nuclear channel image was thresholded to generate a foreground mask. The original image was then filtered and local maxima, representing individual nuclei, were detected and used as seeds for watershed segmentation. (B) RUES2-RFP-SMAD1 grown on micropatterned colonies in E7 and stimulated with BMP4 (50 ng/mL). The RFP-SMAD1 nuclear signal (normalized to H2B) was quantified in individual cells within the micropatterned colony. (C–D) Binary analysis of the SMAD1 response data shown in Figure 2D. (C) The histogram of the RFP-SMAD1 response at T = 14 hr following stimulation with BMP4 was thresholded using Otsu’s method (dashed line). (D) The threshold was applied at each time point to determine the faction of SMAD1 +nuclei as a function of radial position within the colony. Error bars represent the standard deviation over n = 5 colonies from one experiment. (E) RUES2-mCit-SMAD2 grown on micropatterned colonies in E7 stimulated and with ACTIVIN (10 ng/mL). The mCitrine-SMAD2 nuclear-to-cytoplasmic ratio was quantified in individual cells within the micropatterned colony.

Figure 2—video 1. RUES2-RFP-SMAD1 micropatterned colony stimulated with BMP4.

Download video file ^{(13MB, mp4)}

DOI: 10.7554/eLife.38279.007

BMP4 (50 ng/mL, added at T = 0 hr). Images were acquired every 10 min in two channels corresponding to H2B-mCitrine and RFP-SMAD1. Time stamp shows hours and minutes (hh:mm). Colony radius, 500 µm.

Figure 2—video 2. RUES2-RFP-SMAD1 micropatterned colony unstimulated.

Download video file ^{(12.6MB, mp4)}

DOI: 10.7554/eLife.38279.008

Images were acquired every 10 min in two channels corresponding to H2B-mCitrine and RFP-SMAD1. Time stamp shows hours and minutes (hh:mm). Colony radius, 500 µm.

Figure 2—video 3. RUES2-mCit-SMAD2 micropatterned colony stimulated with ACTIVIN.

Download video file ^{(19.5MB, mp4)}

DOI: 10.7554/eLife.38279.009

ACTIVIN (10 ng/mL, added at T = 0 hr). Images were acquired every 10 min in two channels corresponding to H2B-mCherry and mCitrine-SMAD2. Time stamp shows hours and minutes (hh:mm). Colony radius, 500 µm.