RNA visualization of transcriptional bursting in acute and chronically ERα stimulated states.

a. RNA visualization achieved via 24xMS2 insertion into the endogenous NRIP1 loci via CRISPR-Cas9, and selection was improved through the addition of either a fluorescent or puromycin selection marker to the HR donor cassette just upstream of the stop codon. b. An example of a bust event at NRIP1 loci in response to E2 stimulation. c. RNA visualization FISH validation of tagged NRIP1 allele using 24xMS2 showed 94.7% overlap (n=18). d. Long-term timelapse show 3 clear phases of bursting in response to E2 treatment: high bursting post-E2 3 clear phases of bursting in response to E2 treatment: high bursting activity from 24 to 132 minutes post-E2 treatment (Acute); sustained lower bursting activity immediately following Acute phase (Chronic); and low bursting activity post-E2 but preceding Acute phase (pre-Acute). (n=922) e. Acute and chronic bursts of NRIP1 differ significantly in duration, with longer on burst duration during acute phase. f. Acute and chronic bursts of NRIP1 differ significantly in frequency, with longer off-periods in chronic phase. g. PRO-seq of Mega-Trans and ER bound loci under minus, acute, and chronic E2 induction. h. GRO-seq of NRIP1 enhancer and gene expression levels shown on genome browser. i. GRO-seq showing NRIP1 expression levels (FPKM) are ∼1.8x higher in the acute relative to the chronic state. j. Time-course of NRIP1 enhancer and gene transcript levels after E2 stimulation measured by qPCR. (All statistics: 2-Tailed Paired T Test * = p<0.05, ** = p<0.01, *** = p<0.001, **** = p<0.0001)

DNA visualization of loci kinetics in acute versus chronic ERα induced states.

a. DNA visualization via endogenous tagging of NRIP1 by knock-in of 144xCuO using Bxb1 integrase. b. DNA visualization FISH validation of tagged NRIP1 allele using 144xCuO showed 100% overlap. c. E2 minus condition shows significantly lower mobility than stimulated E2 state (n=15). d. E2 minus condition shows significantly lower mobility than either the acute or chronic stimulation states, but there is also a significant further increase in mobility in the chronic versus acute stimulation states. e. Time-course of averaged normalized displacement of NRIP1 under acute versus chronic stimulation. f. Contrails of NRIP1 loci in acute versus chronic states, showing increased displacement under chronic E2 stimulation. (All statistics: 2-Tailed Paired T Test * = p<0.05, ** = p<0.01, *** = p<0.001, **** = p<0.0001)

Biophysical insight into transcription dynamics via single molecule tracking.

a. Live imaging shows that ERα foci are distinct from but in close proximity to both SC35 and the nuclear matrix. b. ERα ChIP-seq shows increased ERα binding genome wide upon E2 induction, where ER is more robustly bound in the acute phase relative to chronic. c. ChIP-seq of MegaTrans factors GATA3, AP2γ, and RARα show similar increased genome wide binding upon E2 induction, with more robust binding in acute versus chronic states. d. Agglomerative clustering shows an acute specific population of molecules (cluster 6, red) with the lowest diffusion coefficient and varies in sub-diffusivity as reflected by the range in alpha. (n=230) e. GMM shows the same acute specific population of molecules (cluster 6, red) with the lowest diffusion coefficient and varies in sub-diffusivity as reflected by the range in alpha. f. Significant enrichment of the lowest diffusion-coefficient, low-subdiffusivity fraction in the acute versus chronic stimulations. g. Analysis of 1,6-hexandiol treatment shows particular disruption of ERα molecules from acute specific cluster 6. h. Bar graphs show fold change of ERα molecules in each GMM cluster when treated with 1,6-hexandiol, with cluster 6 showing the greatest difference in fold change between acute and chronic states. (All statistics: 2-Tailed T Test * = p<0.05, ** = p<0.01, *** = p<0.001, **** = p<0.0001)

DNA RNA dual visualization of mobility relative to transcriptional bursting.

a. Addition of RNA-burst visualization into DNA-visualization NRIP cell line. b. DNA RNA overlap in live cells confirms tagging of same NRIP allele. c. DNA mobility corresponds to gene bursting, where a significant reduction in mobility occurs during burst. (n=10) d. Graph mapping DNA displacement against an RNA burst, showing decrease in displacement during a burst event (characterized in binary form of 0 for no bust and 0.05 for burst signal, trend line is a rolling average of burst signal). (All statistics: 2-Tailed Paired T Test * = p<0.05, ** = p<0.01, *** = p<0.001, **** = p<0.0001)

ERα regulated locus and its association with nuclear sub-structures.

a. Immuno-DNA FISH of NRIP1 shows increased proximity to SC35 in acute phase relative to chronic. (n=822) b. Immuno-DNA FISH of NRIP1 shows induced association with Matrin3 network in the chronic state relative to acute. c. Three-color live visualization of NRIP1 RNA via 24xMS2 and virally transduced SC35 and Matrin3 fused to fluorescent tags. d. Live imaging of NRIP1 relative to SC35 shows peak induced proximity during the acute phase versus chronic. e. Three-color living visualization of NRIP1 DNA, NRIP1 RNA, and endogenously tagged SC35. f. Although E2 stimulation caused induced proximity of NRIP1 to SC35 in general, there is no correlation between proximity and a burst event. (n=10) (All statistics: 2-Tailed Paired T Test * = p<0.05, ** = p<0.01, *** = p<0.001, **** = p<0.0001)

DNA kinetics as a surrogate for detecting condensate formation during burst events.

a. Transcriptional pausing of NRIP1 via flavopiridol treatment in WT MCF7 and NRIP1 DNA visualization cell line measured by qPCR. b. Increased in DNA mobility post flavopiridol treatment detected at NRIP1 loci. (n=68) c. SMT detects a significant depletion of previously identified acute specific, 1,6-hexanidiol sensitive ERα cluster post flavopiridol treatment. d. Bar graphs show fold change of ERα molecules per cluster post flavopiridol, with most confined cluster 6 showing the greatest depletion and less restricted clusters 1 and 3 gaining in ERα molecules. e. Alpha vs diffusion for ERα molecules in E2 minus, acute, and acute plus flavopiridol conditions. f. Flavopiridol treatment results in a statistical gain in alpha for ERα molecules. g. GMM clustering showed statistical significant gains and losses in the most confined (cluster 6) and least restricted (cluster 1) clusters in response to E2 stimulation and flavopiridol induced transcriptional pausing. (All statistics: 2-Tailed Paired T Test * = p<0.05, ** = p<0.01, *** = p<0.001, **** = p<0.0001)

DNA visualization of NRIP1 via CymR protein fused to either mTurquoise2, mKate, or Halo JaneliaFluor-549.

ERα SMT.

a. Agglomerative clustering identified 6 distinct ERα populations characterized by alpha and diffusion in minus E2, acute, and chronic states. MSD curves of a randomized subset of single molecules from each cluster are graphed. b. GMM clustering similarly identified 6 distinct ERα populations characterized by alpha and diffusion in minus E2, acute, and chronic states. MSD curves of a randomized subset of single molecules from each cluster are graphed.

TFF1 FISH.

a. Immuno-DNA FISH of TFF1 shows increased proximity to SC35 in acute phase relative to chronic. b. Immuno-DNA FISH of TFF1 shows induced association with Matrin3 network in the chronic state relative to acute. c. Immuno-RNA FISH of TFF1 shows increased proximity to SC35 in acute relative to chronic phase.