Panel A shows western blot bands comparing alpha-smooth muscle actin levels after transforming growth factor beta treatments at different durations. Panel B shows fluorescence microscopy images of LX-2 cells stained for alpha-smooth muscle actin with and without transforming growth factor beta treatment. Panel C shows a scatter plot quantifying alpha-smooth muscle actin intensity between untreated and transforming growth factor beta-treated cells. The vertical axis represents alpha-smooth muscle actin intensity in arbitrary units, while the horizontal axis represents treatment conditions. Panel D shows fluorescence microscopy images comparing intracellular collagen type 1 alpha 1 localization in cells with or without transforming growth factor beta treatment. Panel E shows a scatter plot quantifying intracellular collagen type 1 alpha 1 integrated density between treatment conditions. The vertical axis represents collagen type 1 alpha 1 intensity per cell, while the horizontal axis represents treatment conditions. Panel F shows fluorescence microscopy images displaying colocalization between Col1A1-mNeonGreen and anti-Col1A1 staining inside cells and extracellular fibrils. Panel G shows fluorescence microscopy images and zoomed insets illustrating Col1A1 condensate colocalization patterns within intracellular regions. Panel H shows a scatter plot comparing Manders colocalization coefficients M1 and M2 for Col1A1 channels. The vertical axis represents colocalization coefficient values, while the horizontal axis represents M1 and M2 categories. Panel I shows fluorescence microscopy images comparing endogenous Col1A1 distribution with ss-KDEL-mScarlet endoplasmic reticulum labeling. Panel J shows a scatter plot comparing coefficients of variation for Col1A1 and ss-KDEL-mScarlet fluorescence distributions. The vertical axis represents coefficient of variation values, while the horizontal axis represents fluorescence marker categories. Panel K shows fluorescence microscopy images of endogenous Col1A1 patches and condensate-like structures within U2OS cells. Panel L shows a protein domain schematic and disorder prediction graph for Col1A1 across amino acid positions. The horizontal axis represents amino acid position, while the vertical axis represents disorder score values. Panel M shows a structural model of a heterotrimeric procollagen complex with labeled residues and measured distance annotations.
PC1 is intrinsically predisposed for LLPS. (A) Western blot showing an increase in ⍺-SMA levels in LX-2 cells upon treatment with 10 or 20 ng/ml of TGF-β for 24 and 48 h. Calnexin levels are used as loading controls. Throughout the results, +TGF-β refers to the condition indicated by the red arrowhead. (B) Fluorescence images of fixed LX-2 cells grown with or without TGF-β and stained using antibodies against ⍺-SMA (scale bar: 20 µm). (C) Plot quantifying mean ⍺-SMA intensity in cells grown with or without TGF-β. A total of 26 cells grown without TGF-β and 27 cells grown with TGF-β were analyzed (****P < 0.0001). (D) Fluorescence images of fixed LX-2 cells grown in the presence or absence of TGF-β. Images of cells grown without TGF-β with brightness and contrast equal to the +TGF-β condition (middle) and the same image with brightness and contrast enhanced (right) (scale bars: 20 µm). For comparison, an expanded version of Fig. 1 G is used as the +TGF-β image. (E) Plot quantifying fluorescence integrated density of intracellular COL1A1 in LX-2 cells grown with or without TGF-β. A total of 63 cells for each case from two independent experiments were analyzed (****P < 0.0001). (F) Fluorescence images of fixed LX-2 cells stably expressing Col1A1-mNG stained with antibodies against Col1A1 (top) and zoomed inset (bottom). Arrowheads indicate colocalization between Col1A1-mNG and anti-Col1A1 both inside the cell and in fibrils formed outside (scale bars: 10 µm; inset: 5 µm). (G) Fluorescence images of fixed LX-2 cells stably expressing Col1A1-mNG stained with antibodies against Col1A1 (top) and zoomed inset (bottom) showing colocalization between the two channels in Col1A1 condensates (scale bars: 10 µm; inset: 5 µm). (H) Manders’ colocalization coefficients M1 and M2 for images of LX-2 cells stably expressing Col1A1-mNG and stained with anti-Col1A1. A total of 13 ROIs from two independent experiments were analyzed (ns: not significant). (I) Fluorescence images of LX-2 cells grown in the absence of TGF-β, transfected with ss-KDEL-mScarlet, fixed, and stained for endogenous Col1A1 (top), and zoomed insets (bottom). White arrowheads in the Col1A1 channel indicate nonhomogeneous patches of Col1A1 and clusters resembling condensates, and those in the merged channel indicate patches of the ER that are not enriched in Col1A1 (scale bars: 10 µm; inset: 2 µm). (J) Plot comparing the CV in Col1A1 and ss-KDEL-mScarlet fluorescence across the same ROIs spanning the ER. A total of 59 ROIs from 20 cells from two independent experiments were analyzed (****P < 0.0001). (K) Fluorescence images of fixed U2OS cells stained for endogenous Col1A1 (left) and zoomed inset (right). White arrowheads indicate nonhomogeneous patches of Col1A1 and clusters resembling condensates (scale bars: 10 µm; inset: 2 µm). (L) Domain organization of Col1A1 and disorder scores across the protein predicted using IUPred2A (Erdős and Dosztányi, 2020). (M) Structure of a heterotrimer of PC1 with two ⍺1 chains and one ⍺2 chain predicted using the AlphaFold3 server (Abramson et al., 2024). The distance between residues P1011 and L1201 is indicated by the dashed line. ss-KDEL-mScarlet, signal-sequence-KDEL-mScarlet. Source data are available for this figure: SourceData FS1.