Figure 3. Proteins that aggregate after topoisomerase I poisoning are supersaturated and prone to engage in liquid-liquid phase separation (LLPS).

See also Figure 3—figure supplements 1 and 2. (A) TANGO scores of HEK293T whole-cell lysate (WCL), nonaggregated proteins (NIA), and aggregated fractions. (B) Protein abundance of HEK293T WCL, nonaggregated proteins (NIA), and aggregated fractions as measured by label-free quantification (LFQ) intensities. (C) Transcript abundances of HEK293T WCL, nonaggregated proteins (NIA), and aggregated fractions (as measured by RNAseq). (D) Supersaturation scores of HEK293T WCL, nonaggregated proteins (NIA), and aggregated fractions. (E) Clarification of (F). (F) Transcript abundances (as measured by RNAseq) plotted against TANGO scores for the complete HEK293T MS/MS analysis. Proteins above the diagonal (=HEK293T median saturation score, calculated using the HEK293T WCL dataset) are relatively supersaturated. (G) catGRANULE scores for the indicated protein fractions in HEK293T cells. In all graphs, individual proteins and median values (red lines) are shown. p-Values are obtained by Kruskal–Wallis tests followed by Dunn’s correction for multiple comparisons.

Figure 3—figure supplement 1. Proteins that aggregate after camptothecin treatment and ataxia telangiectasia mutated (ATM) loss represent a vulnerable subfraction of the proteome.

(A) CamSol-intrinsic (in)solubility scores of complete whole-cell lysate (WCL), nonaggregated proteins (NIA), and aggregated fractions in HEK293T cells. Dotted line indicates the theoretical threshold of relative intrinsic (in)solubility. See text for reference. (B) Protein abundances of HEK293T WCL and aggregated fractions obtained by cross-referencing with a HEK293 NSAF (normalized spectral abundance factor) reference proteome, as well as the protein abundances of the entire HEK293 reference proteome itself (see text for reference). (C) Supersaturation scores obtained by cross-referencing with the human composite supersaturation database generated by Ciryam et al. (see text for reference) for HEK293T cells. (D) PScores of complete WCL, nonaggregated proteins (NIA), and aggregated fractions in HEK293T. (E) Presence of proteins with a high PScore (> 4) in the indicated fractions in HEK293T. See also text. (F) Distribution of supersaturated (Sup) and non-supersaturated (NonSup) proteins in the HEK293T and U2OS NIA fractions. (G). catGRANULE scores of supersaturated (Sup) and non-supersaturated (NonSup) proteins in HEK293T and U2OS NIA fractions. (H) TANGO scores of complete WCL, nonaggregated proteins (NIA), and aggregated fractions in U2OS cells. (I) CamSol-intrinsic (in)solubility scores of complete WCL, nonaggregated proteins (NIA), and aggregated fractions in U2OS cells. Dotted line indicates the theoretical threshold of relative intrinsic (in)solubility. (J) Transcript abundances (as measured by RNAseq) of the indicated fractions in U2OS cells. (K) Protein abundance as measured by label-free quantification (LFQ) intensities of the indicated fractions in U2OS cells. (L) Protein abundances of U2OS WCL and aggregated fractions obtained by cross-referencing with a U2OS normalized spectral abundance factor (NSAF) reference proteome, as well as the protein abundances of the entire U2OS reference proteome itself (see text for reference). (M) Supersaturation scores for the indicated protein fractions in U2OS cells. (N) Supersaturation scores obtained by cross-referencing with the supersaturation database generated by Ciryam et al. (see text for reference) for U2OS cells. (O) Transcript abundances (as measured by RNAseq) plotted against TANGO scores for the complete U2OS MS analysis. All proteins above the diagonal (=U2 OS median saturation score, calculated using the U2OS WCL dataset) are relatively supersaturated. See also Figure 3E and F. (P) catGRANULE scores for the indicated protein fractions in U2OS cells. (Q) PScores of complete WCL, nonaggregated proteins (NIA), and aggregated fractions in U2OS. (R) Presence of proteins with a high PScore (>4) in the indicated fractions in U2OS. For all bar graphs, bars represent mean ± SEM. For all other graphs, circles represent individual proteins, red lines indicate the median value. p-Values are obtained by Kruskal–Wallis tests followed by Dunn’s correction for multiple comparisons, except in N, where two-tailed Mann–Whitney tests were used.

Figure 3—figure supplement 2. GO term analyses of RNAseq data.

GO term analysis of the differentially expressed genes (–1 > log₂FC > 1) RNAseq data of the indicated cell line after the indicated treatment. Upper panels: upregulated processes are shown. Lower panels: downregulated processes are shown. Note that HEK293T cells treated with ataxia telangiectasia mutated (ATM) inhibitor are not included as only two genes were differentially expressed (see also Supplementary file 2).