Supplemental Material

This article contains the following supporting material:

• Figure S1 - Common sequence and intrinsic disorder features revealed by the initial mapping of VHL. (A and B) Schematic representation of VHL segments shown in (Figure 1B) are shown here in (A), with domains of interest represented using the same color code as in (B). Images (top of (A)) used for scoring of nucleolar localization are duplicated from Figure 1B for presentation purposes. Insets show Hoechst staining of DNA and scale bars represent 10 μm. (B) Amino acid sequence of VHL, depicting arginine-rich or hydrophobic segments in different colours. (C) Computational analysis of VHL using the disEMBL predictor of disorder. Arginine-rich segments and hydrophobic repeats are presented below the graph following the same color code as in (A) and (B). The arginine domains are located at regions of disorder probabilities of less than 0.03, 0.2, and 0.32 compared to threshold levels for disorder (dotted lines) of 0.09, 0.5, and 0.43, following the predictors Hot-Loops, Remark-465, and Loops or Coil.
• Figure S2 - Fluorescence loss in photobleaching (FLIP) analysis of VHL's subnuclear targeting sequences and control proteins. MCF7 cells were transfected to express low levels of the indicated GFP-tagged proteins or fragments. Cells were then submitted to hypoxic treatment under SD (A) or AP (B-E) conditions and FLIP analysis was performed. Bleached nuclear regions (squares) are shown and pseudocolored panels are included to better illustrate subtle changes in fluorescence intensity.
• Figure S3 - Prediction of effect of mutations on disorder level. Effects of spacing between arginine residues and hydrophobic repeats within the fusion construct 107-113-HD1 on the level of the intrinsic disorder of STAD are illustrated.
• Figure S4 - R-R-I/L-X₃-R domains but not hydrophobic repeats are positioned within regions exhibiting low structural disorder. Examples of computational analysis of VHL (A) and RNF8 (B) using the disEMBL predictor of disorder are shown. Arginine-rich segments (gray) and hydrophobic repeats (black) are presented by boxes on the x axis. Arginine-rich domains are located at regions with below-threshold disorder probabilities as revealed by the predictors Hot-Loops, Remark-465, and Loops or Coil.
• Figure S5 - Confirmation of predicted changes to NoDS^H+-containing proteins. (A-C) Assessment of VHL and B23 control kinetics by FRAP analysis. MCF7 cells, transiently transfected to express low levels of VHL-GFP (A, B) or B23-GFP (C), were incubated under indicated conditions and imaged before and after bleaching of the square-marked regions. Post-bleach time is indicated in seconds and pseudocolored panels are included to better illustrate minimal changes in fluorescence. Fluorescence recovery value R is shown. (D) FRAP analysis of nucleolar RNF8 reveals its static detention within the nucleolar space. MCF7 cells expressing low levels of GFP-tagged RNF8 were incubated under hypoxia (1%O₂) in AP conditions. A region corresponding to about one half of a nucleolus was bleached and cells were monitored by time-lapse fluorescence microscopy. Scale bar represents 1 μm. (E) Interaction with the nucleolus is required for static detention. MCF7 cells expressing high levels of GFP-tagged RNF8 were incubated under hypoxia (1%O₂) in AP conditions then submitted to FLIP analysis. An overlay of GFP fluorescence before bleaching (shown in green) and at endpoint (shown in red) is shown to ensure that the relative position of nucleoli within the nucleus did not change over the course of the experiment. Scale bar represents 5 μm. (F) Acidosis-dependent nucleolar detention of RNF8 is reversible. Following hypoxia-induced acidification of AP media (pH 6.3) and confinement of RNF8-GFP to the nucleolus, cells were either immediately submitted to nucleolar FRAP analysis (blue, pH 6.3) or replenished with fresh SD media to induce the reversion of RNF8-GFP to the nucleo-cytoplasm first then submitted to FRAP analysis in which small nucleoplasmic regions were bleached (black, pH 6.3 to 7.2). Results reveal that RNF8-GFP resumes its high mobility profile following a return to neutral pH conditions. (G) pH-dependent nucleolar detention of RNF8 affects its participation in nucleoplasmic molecular networks. MCF7 cells, unaltered or transfected to express high levels of RNF8-GFP, were incubated under neutral or acidotic conditions. Cells were then harvested and the level of CRBPII transcripts was detected by semi-quantitative reverse-transcriptase PCR. Transcript levels were quanified using Actin as control.
• Figure S6 - Both the steady-state distribution and dynamic profile of the non-NoDS^H+-containing HSP110 are insensitive to changes in extracellular pH. MCF7 cells expressing low levels of HSP110-GFP were incubated under SD (A) or AP (B) conditions and imaged before and after bleaching the square-marked regions. Post-bleach time is indicated in seconds and pseudocolored panels and arrows are included to better illustrate minimal changes in fluorescence. Fluorescence recovery value R is shown.
• Figure S7 - R-R-I/L domains but not hydrophobic repeats are positioned within regions exhibiting low structural disorder. Examples of computational analysis of HSC70 (A) and PBK1 (B) using the disEMBL predictor of disorder are shown. Arginine-rich domains are located at regions with below-threshold disorder probabilities as revealed by the predictors Hot-Loops, Remark-465, and Loops or Coil.
• Table S1 - List of candidate proteins predicted to exhibit pH-dependent nucleolar detention in response to acidosis based on the STAD sequence R-R-I/L-X₃-R. Results of a search of the UniProtKB/SwissProt database using motifs R-R-I/L-X₃-R-X_0-100-L-θ-V/L and L-θ-V/L-X_0-100-R-R-I/L-X₃-R, where θ symbolizes any hydrophobic residue. STADs were filtered for low disorder using the disEMBL software. Nu, nucleus; Cyto, cytoplasm; PML-NBs, PML nuclear bodies; ER, endoplasmic reticulum.
• Table S2 - Correlation between conformity to the rules and nucleolar targeting activity.
• Table S3 - List of candidate proteins predicted to exhibit pH-dependent nucleolar detention in response to acidosis based on the STAD sequence R-R-I/L. Results of a search of the UniProtKB/SwissProt database using motifs R-R-I/L-X_0-110-L-θ/N-V/L and L-θ/N-V/L-X_0-110-R-R-I/L, where θ symbolizes any hydrophobic residue. STADs were filtered for low disorder using the disEMBL algorithm. Shown is a partial list of proteins conforming to all three rules. Cyto, cytoplasm; Nu, nucleus; Nol, nucleolus; NA, unknown.