Journal: bioRxiv

Article Title: Multiple ubiquitin ligases protect human genome integrity by targeting cancer-associated APOBEC3 deaminases for degradation

doi: 10.1101/2024.04.23.590688

Figure Lengend Snippet: (a) A monoclonal RKO cell line harboring a DOX-inducible Cas9-P2A-BFP expression construct, and a constitutively expressed mCherry-A3H-II-P2A-EGFP-A3H-I dual- reporter (RKO-DOX-Cas9-dualA3H) was established. Unstable EGFP-A3H-I and stable- mCherry-A3H-II are synthesized in equimolar amounts, yet EGFP-A3H-I accumulates at low steady-state levels resulting from its proteasomal degradation. (b) Schematic representation of CRISPR/Cas9 FACS-based screening procedure. RKO-DOX-Cas9-dualA3H cells were transduced with a ubiquitin-focused sgRNA library. After selection of sgRNA-positive cells with G418, Cas9 expression was induced with DOX for 3 and 6 days, followed by sorting of cells with the 1-2% highest and lowest EGFP or mCherry fluorescence by flow cytometry. Integrated sgRNA sequences were identified by next generation sequencing (NGS), and their differential abundance relative to non-sorted cells analyzed. (c) Targeted genes enriched in EGFP-A3H- I high sorted cell populations 6 days post Cas9 induction. (d) Heatmap of top genes on log2 fold- change and p-value grouped by functional categories. Genes enriched in EGFP-A3H-I high cell populations 6 days post Cas9 induction with a log2 fold-change >0.6 which were not enriched in mCherry high or GFP low on either day 3 (LFC > 0.45) or day 6 (LFC > 0.6). Adjusted p-values are based on MaGECK analysis of three independent replicate sorts. Dashed lines indicate a log2 fold-change < 0.6. E3 (E3 ligases), E2 (E2 conjungating enzymes), DUB (Deubiquitinases), CUL (components of Cullin-RING ubiquitin ligase complex). (e) Monoclonal RKO-Dox-Cas9-dualA3H cells were transduced with individual sgRNAs targeting the indicated genes. Subsequently, Cas9 expression was induced by DOX for 6 days and EGFP-A3H-I and mCherry-A3H-II mean fluorescence intensity (MFI) analyzed by flow cytometry, and (f) quantified (means and SD, one-way ANOVA, **** p < 0.0001, *** p < 0.0005, n = 3). (g) RKO cells harboring DOX-inducible Cas9 were transduced with sgRNAs targeting UBR4 , UBR5 , or HUWE1 individually, or combined, and sorted for sgRNA-positive cells. Gene editing was induced with DOX for 6 days, after which endogenous A3B protein levels were determined by WB, and (h) quantified (means and SD, one-way ANOVA, **** p < 0.0001, ** p < 0.005, * p < 0.05, n = 2).

Article Snippet: Membranes were blocked in 5% BSA in PBS-T for 1 h. at RT, and subsequently incubated with primary antibodies diluted in 5% BSA overnight at 4 °C (ARP10 Antibody (Novus, 1:1000), Anti-APOBEC3B Antibody (Abcam, 1:1000), Anti-MYC antibody (Sigma-Aldrich, 1:5000), HA-Tag (C29F4) Rabbit mAb (Cell Signaling Technology, 1:1000), HA-Tag (6E2) Mouse mAb (Cell Signaling Technology, 1:1000), OLLAS Epitope Tag Antibody (L2) (Novus, 1:4000), LC3B Antibody ((Cell Signaling Technology, 1:1000), Ubiquitin (P4D1) (Santa Cruz Biotechnology, 1:1000), Anti-UBR4/p600 antibody (Abcam, 1:1000), Rabbit anti-EDD1 Antibody (Bethyl, 1:1000), Rabbit anti-Lasu1/Ureb1 Antibody (HUWE1) (Bethyl, 1:1000), Monoclonal Anti-α-Tubulin antibody produced in mouse (Sigma-Aldrich, 1:1000), Lamin A/C Antibody (E-1) (Santa Cruz Biotechnology, 1:1000), Monoclonal Anti- Vinculin antibody (Sigma-Aldrich, 1:1000), Anti-beta Actin antibody (HRP) (Abcam, 1:20000)).

Techniques: Expressing, Construct, Synthesized, CRISPR, Transduction, Selection, Fluorescence, Flow Cytometry, Next-Generation Sequencing, Functional Assay

Journal: bioRxiv

Article Title: Multiple ubiquitin ligases protect human genome integrity by targeting cancer-associated APOBEC3 deaminases for degradation

doi: 10.1101/2024.04.23.590688

Figure Lengend Snippet: (a) Monoclonal RKO-DOX-Cas9-dualA3H cells were treated with CHX or EPOX for 5 h., followed by analysis of EGFP-A3H-I or mCherry-A3H-II levels by flow cytometry. (b) Targeted genes enriched in EGFP-A3H-I high sorted cell populations 3 days post Cas9 induction. (c) Heatmap of top genes on log 2 fold-change and p-value grouped by functional categories. Genes enriched in EGFP-A3H-I high cell populations 3 days post Cas9 induction with a log 2 fold-change > 0.45 which were not enriched in mCherry high or GFP low on the same day. Dashed lines indicate a log 2 fold-change < 0.45. Adjusted p-values are based on MaGECK analysis of three independent replicate sorts. E3 (E3 ligases), E2 (E2 conjungating enzymes). (d) Polyclonal RKO-DOX-Cas9-dualA3H cells were transduced with sgRNA vectors targeting UBR4 , UBR5 , or HUWE1 . EGFP-A3H-I and mCherry-A3H-II abundance was analyzed by flow cytometry 6 days after Cas9 induction. The mean fluorescence intensity (MFI) was quantified (one-way ANOVA compared to sg AAVS1 , ** p < 0.005, * p < 0.05, n = 3). (e) RKO-DOX-Cas9-MYC-mCherry-P2A-3xHA-A3H-I cells were transduced with sgRNA simultaneously targeting either the three E3 ligases or control loci ( AAVS1 , CCR5 , hROSA ), gene editing induced with DOX for 6 days and then treated with CHX for different times. A3H-I protein levels were determined by WB and (f) the half-life quantified from single-step exponential decay curves, statistics were calculated by two-way ANOVA, * p < 0.05 (n = 3). (g) THP-1 cells harboring DOX-inducible Cas9 were transduced with sgRNAs targeting UBR4 , UBR5 , or HUWE1 , and sorted for sgRNA-positive cells. Gene editing was induced with DOX for 3 days, after which endogenous A3B and A3G protein levels were determined by WB, and (h) quantified. Data represent single experiment. (i) RKO cells harboring DOX-inducible Cas9 were transduced with sgRNAs targeting UBR4 , UBR5 , or HUWE1 , and sorted for sgRNA-positive cells. Gene editing was induced with DOX for 6 days, after which endogenous A3B mRNA levels were determined by RT-qPCR (means and SD, one-way ANOVA, ns p > 0.1, n = 3).

Article Snippet: Membranes were blocked in 5% BSA in PBS-T for 1 h. at RT, and subsequently incubated with primary antibodies diluted in 5% BSA overnight at 4 °C (ARP10 Antibody (Novus, 1:1000), Anti-APOBEC3B Antibody (Abcam, 1:1000), Anti-MYC antibody (Sigma-Aldrich, 1:5000), HA-Tag (C29F4) Rabbit mAb (Cell Signaling Technology, 1:1000), HA-Tag (6E2) Mouse mAb (Cell Signaling Technology, 1:1000), OLLAS Epitope Tag Antibody (L2) (Novus, 1:4000), LC3B Antibody ((Cell Signaling Technology, 1:1000), Ubiquitin (P4D1) (Santa Cruz Biotechnology, 1:1000), Anti-UBR4/p600 antibody (Abcam, 1:1000), Rabbit anti-EDD1 Antibody (Bethyl, 1:1000), Rabbit anti-Lasu1/Ureb1 Antibody (HUWE1) (Bethyl, 1:1000), Monoclonal Anti-α-Tubulin antibody produced in mouse (Sigma-Aldrich, 1:1000), Lamin A/C Antibody (E-1) (Santa Cruz Biotechnology, 1:1000), Monoclonal Anti- Vinculin antibody (Sigma-Aldrich, 1:1000), Anti-beta Actin antibody (HRP) (Abcam, 1:20000)).

Techniques: Flow Cytometry, Functional Assay, Transduction, Fluorescence, Quantitative RT-PCR

Journal: bioRxiv

Article Title: Multiple ubiquitin ligases protect human genome integrity by targeting cancer-associated APOBEC3 deaminases for degradation

doi: 10.1101/2024.04.23.590688

Figure Lengend Snippet: (a) 10x-His-MBP-A3H-I, 10x-His-MBP-A3H-II and 10xHis-MBP-A3H-II-RBM (E56A/W115A/ R175E/ R176E) were expressed in E. coli , purified by HisTrap and subsequent gel filtration, after which purified proteins were analyzed by PAGE and Coomassie staining. b) Purified proteins were analyzed by denaturing Urea-TBE PAGE followed by SYBR Gold staining. (c-e) In vitro ubiquitination assays were performed with recombinant (c) UBR5, (d) HUWE1, or (e) UBR4 and WT A3H-I/II or A3H-RBM as substrates, and in the presence of DyLight488-labeled recombinant ubiquitin. Subsequently, A3H was immunoprecipitated using anti-MBP-coupled beads and the ubiquitination pattern visualized by fluorescent imaging for DyLight488. f-h) In vitro ubiquitination assay with recombinant f) UBR5, g) HUWE1, h) UBR4 and A3H-I/II WT or A3H-RBM in the absence or presence of RNAse A. Ubiquitinated A3H was visualized as in (c-e).

Article Snippet: Membranes were blocked in 5% BSA in PBS-T for 1 h. at RT, and subsequently incubated with primary antibodies diluted in 5% BSA overnight at 4 °C (ARP10 Antibody (Novus, 1:1000), Anti-APOBEC3B Antibody (Abcam, 1:1000), Anti-MYC antibody (Sigma-Aldrich, 1:5000), HA-Tag (C29F4) Rabbit mAb (Cell Signaling Technology, 1:1000), HA-Tag (6E2) Mouse mAb (Cell Signaling Technology, 1:1000), OLLAS Epitope Tag Antibody (L2) (Novus, 1:4000), LC3B Antibody ((Cell Signaling Technology, 1:1000), Ubiquitin (P4D1) (Santa Cruz Biotechnology, 1:1000), Anti-UBR4/p600 antibody (Abcam, 1:1000), Rabbit anti-EDD1 Antibody (Bethyl, 1:1000), Rabbit anti-Lasu1/Ureb1 Antibody (HUWE1) (Bethyl, 1:1000), Monoclonal Anti-α-Tubulin antibody produced in mouse (Sigma-Aldrich, 1:1000), Lamin A/C Antibody (E-1) (Santa Cruz Biotechnology, 1:1000), Monoclonal Anti- Vinculin antibody (Sigma-Aldrich, 1:1000), Anti-beta Actin antibody (HRP) (Abcam, 1:20000)).

Techniques: Purification, Filtration, Staining, In Vitro, Recombinant, Labeling, Immunoprecipitation, Imaging, Ubiquitin Assay

Journal: bioRxiv

Article Title: Multiple ubiquitin ligases protect human genome integrity by targeting cancer-associated APOBEC3 deaminases for degradation

doi: 10.1101/2024.04.23.590688

Figure Lengend Snippet: (a) Cell lysates from a UNG2 -/- RKO cell line expressing DOX-inducible Cas9 and mCherry-P2A-3xHA- A3H-I were analyzed by WB for transgene expression. (b) These cells we transduced with sgRNAs targeting UBR4 , UBR5 , or HUWE1 , and subsequently sorted for sgRNA-positive cells. Gene editing was induced with DOX for up to four cell doublings, after which the protein levels of the targeted E3 ligases were analyzed by WB. (c) Best subset signature refitting of mutREAD data, averaged per genotype, using signatures related to overactivity of APOBEC family enzymes and signatures commonly active in colon carcinoma, the parent tumor type for the model cell line RKO. For the purpose of comparing signature activity across genotypes, samples were matched by doubling time. Each bar represents the signature refitting results, scaled to 1 to show the relative signature contribution. (d) Signature presence test for the indicated signatures. Genotypes (n = 3, a-c), are divided by vertical dashed lines. The y-axis shows the log of the likelihood ratio, representing the maximum likelihood of the data, given a refit model including the signature of interest over the maximum likelihood of the data, given a refit model excluding the signature of interest. A log(likelihood ratio) greater than 0 indicates a significant activity of the signature of interest. The translucence of the bars (-log(p)) indicates the level of significance of the likelihood ratio test compared to the null hypothesis, which assumes that the mutational profile can be reasonably reconstructed without the signature of interest. (e-k) Cancers from TCGA/ICGC were grouped, based on whether the indicated E3 ligase genes were wild-type (wt) or mutated (mut), and the (e) total number of mutations per group plotted. (f-k) Subsequently, mutational signatures were normalized to the total number of mutations in each sample, and the levels of indicated signatures compared between two groups. “E3s comb ” compares all samples, in which UBR4 , UBR5 and HUWE1 are either all wild- type (wt), or at least one of the E3 ligase genes was mutated (mut). Data represent Wilcox rank sum test between wt and mut of each genotype, ns: p > 0.05, *: p ≤ 0.05, **: p ≤ 0.01, ***: p ≤ 0.001, ****: p ≤ 0.0001, n = 2703 cancer genome samples).

Article Snippet: Membranes were blocked in 5% BSA in PBS-T for 1 h. at RT, and subsequently incubated with primary antibodies diluted in 5% BSA overnight at 4 °C (ARP10 Antibody (Novus, 1:1000), Anti-APOBEC3B Antibody (Abcam, 1:1000), Anti-MYC antibody (Sigma-Aldrich, 1:5000), HA-Tag (C29F4) Rabbit mAb (Cell Signaling Technology, 1:1000), HA-Tag (6E2) Mouse mAb (Cell Signaling Technology, 1:1000), OLLAS Epitope Tag Antibody (L2) (Novus, 1:4000), LC3B Antibody ((Cell Signaling Technology, 1:1000), Ubiquitin (P4D1) (Santa Cruz Biotechnology, 1:1000), Anti-UBR4/p600 antibody (Abcam, 1:1000), Rabbit anti-EDD1 Antibody (Bethyl, 1:1000), Rabbit anti-Lasu1/Ureb1 Antibody (HUWE1) (Bethyl, 1:1000), Monoclonal Anti-α-Tubulin antibody produced in mouse (Sigma-Aldrich, 1:1000), Lamin A/C Antibody (E-1) (Santa Cruz Biotechnology, 1:1000), Monoclonal Anti- Vinculin antibody (Sigma-Aldrich, 1:1000), Anti-beta Actin antibody (HRP) (Abcam, 1:20000)).

Techniques: Expressing, Transduction, Activity Assay

Journal: bioRxiv

Article Title: Multiple ubiquitin ligases protect human genome integrity by targeting cancer-associated APOBEC3 deaminases for degradation

doi: 10.1101/2024.04.23.590688

Figure Lengend Snippet: (a) Schematic representation of mutREAD sequencing to identify APOBEC signature mutations in cells. A UNG2 -/- RKO cell line expressing DOX-inducible Cas9 and mCherry-P2A- 3xHA-A3H-I was transduced with sgRNAs targeting UBR4 , UBR5 , or HUWE1 , and subsequently sorted for sgRNA-positive cells. Gene editing was induced with DOX for up to four cell doublings, after which gDNA was isolated and samples prepared for mutREAD sequencing. (b-c) Best subset signature refitting of mutREAD data, averaged per genotype, using signatures related to overactivity of APOBEC family enzymes and signatures commonly active in colon carcinoma, the parent tumor type for the model cell line RKO. For the purpose of comparing signature activity across genotypes, samples were matched by recorded doubling time. Each bar represents the average signature refitting results for 3 technical replicates, (b) scaled to 1 to show the relative signature contribution, or (c) unscaled to show the absolute signature contribution in terms of number of mutations for individual replicates. (d) Schematic representation of mutational signature analysis from PCAWG dataset. (e) Cancers from TCGA/ICGC were grouped, based on whether the indicated E3 ligase genes were wild-type (wt) or mutated (mut). Mutational signatures were normalized to the total number of mutations in each sample. The level of SBS13 APOBEC signature was compared between the two groups. “E3s comb ” compares all samples, in which UBR4 , UBR5 , and HUWE1 are either all wild-type, or at least one of the E3 ligase genes was mutated. Data represent Wilcox rank sum test between wt and mut of each genotype, ns: p > 0.05, *: p ≤ 0.05, **: p ≤ 0.01, ***: p ≤ 0.001, ****: p ≤ 0.0001, n = 2703 cancer genome samples). (f) Model of A3 protein abundance regulation. RNA binding shields A3 proteins from UBR5/HUWE1 binding and degradation, ensuring high antiviral A3 protein levels in the cytosol. Lack of RNA binding enables UBR5/HUWE1 binding to the unengaged interaction interface (red), resulting in A3 degradation. Since unengaged A3s localize to the nucleus, this ensures low nuclear A3 levels, thereby protecting from genomic DNA mutagenesis. UBR4 extends poly-ubiquitin chains in an E4 ligase capacity.

Article Snippet: Membranes were blocked in 5% BSA in PBS-T for 1 h. at RT, and subsequently incubated with primary antibodies diluted in 5% BSA overnight at 4 °C (ARP10 Antibody (Novus, 1:1000), Anti-APOBEC3B Antibody (Abcam, 1:1000), Anti-MYC antibody (Sigma-Aldrich, 1:5000), HA-Tag (C29F4) Rabbit mAb (Cell Signaling Technology, 1:1000), HA-Tag (6E2) Mouse mAb (Cell Signaling Technology, 1:1000), OLLAS Epitope Tag Antibody (L2) (Novus, 1:4000), LC3B Antibody ((Cell Signaling Technology, 1:1000), Ubiquitin (P4D1) (Santa Cruz Biotechnology, 1:1000), Anti-UBR4/p600 antibody (Abcam, 1:1000), Rabbit anti-EDD1 Antibody (Bethyl, 1:1000), Rabbit anti-Lasu1/Ureb1 Antibody (HUWE1) (Bethyl, 1:1000), Monoclonal Anti-α-Tubulin antibody produced in mouse (Sigma-Aldrich, 1:1000), Lamin A/C Antibody (E-1) (Santa Cruz Biotechnology, 1:1000), Monoclonal Anti- Vinculin antibody (Sigma-Aldrich, 1:1000), Anti-beta Actin antibody (HRP) (Abcam, 1:20000)).

Techniques: Sequencing, Expressing, Transduction, Isolation, Activity Assay, RNA Binding Assay, Binding Assay, Mutagenesis