Fig 2: Unified working model: role of NER proteins in 8-oxoguanine repair.Treatment of cells expressing FAP-TRF1 with dye (100 nM, 15 min) plus light (660 nm, 10 min) introduces 8-oxoG lesions at telomeres. In the DDB2-dependent repair pathway, DDB2 recognizes 8-oxoG lesions and facilitates chromatin relaxation through chromatin decompaction allowing the recruitment of XPC and OGG1 to the damage site. OGG1 recruitment facilitates the dissociation of DDB2. In the absence of downstream repair, DDB2 is retained longer at 8-oxoG sites requiring DDB1-Cul4A-RBX1 (CRL) mediated DDB2 dissociation. At actively transcribed regions, OGG1 can access the lesion independent of DDB2. 8-oxoG processing can lead to toxic BER intermediates that can act as a transcription block. Transcription-coupled repair (TCR) proteins, including XPA, participate in the repair of these BER intermediates. (See also Supplementary Movie 3).

Fig 3: When dissociated from DDB1, CRBN and its ID-associated mutants are subjected to SCFFbxo7-mediated destruction.Data are representative of experimental duplicates except for Fig 3A. (A) Western blot analysis for CRBN levels in protein extracts prepared from mouse brain (Ddb1F/F n = 3; Ddb1F/F;Camk2a-Cre n = 4) or hippocampus (Ddb1F/F n = 2; Ddb1F/F;Camk2a-Cre n = 3). Protein signals were measured by NIH Image J. Quantification of CRBN were standardized as the ratio of CRBN signals to the cognate actin signals, followed by normalization to the mean value of standardized CRBN signals in WT mouse brain or hippocampus. (B) Western blot analysis for CRBN levels in indicated cells transfected with siRNA for DDB1 or siRNA for GFP. siGFP, as a non-targeting control. Quantification of CRBN was normalized to the cognate actin signals. (C) Co-immunoprecipitation (co-IP) of HA-tagged WT or mutant CRBN with DDB1 in 293T cells. Whole-cell extracts from 293T cells over-expressing HA-CRBN constructs were immunoprecipitated with anti-HA affinity beads and analyzed by immunoblotting with indicated antibodies. Quantification of DDB1 pulled down by HA-CRBN in IP samples was normalized to the cognate DDB1 signals in input samples. (D) Identification of protein binding region of CRBN with DDB1 by immunoprecipitation of HA-tagged WT, R419X mutant or truncated CRBN with DDB1 after they were over-expressed in 293T cells. Quantification of DDB1 pulled down by HA-CRBN in IP samples was normalized to the cognate DDB1 signals in input samples. (E) Co-immunoprecipitation of CRBNR419X with Fbxo7, Skp1, Cul1 and ubiquitin in 293T cells over-expressing HA-CRBNR419X, Flag-Fbxo7, Myc-Skp1 and Myc-Cul1. (F) Ubiquitination assay for CRBNR419X by immunoprecipitation of CRBNR419X with ubiquitin in 293T cells. 293T cells over-expressing HA- CRBNR419X were transfected with siRNA for Cul1 or control and harvested for IP after treatment with MG132 (10 µM) for 6 hrs. Quantification of Ubiquitin pulled down by HA- CRBNR419X was normalized to ß-actin. (G) Western blot analysis for CRBNR419X levels in 293T cells transfected with HA-CRBNR419X along with siRNA for Ctrl or Cul1. L.E., long exposure. S.E., short exposure. Quantification of CRBNR419X (L.E.) was normalized to ß-actin.

Fig 4: Deficiency of CRL4CRBN promotes the destruction of BK channel a subunit.Data represent at least two independent experimental replicates. (A) Quantification of Slo1 level (detected by anti-Slo1, Novus) in brain extracts isolated from WT and CrbnF/F;Nestin-Cre mice (n = 15 for each group). Protein levels were standardized as the ratio of Slo1 signals to the cognate ß-actin signals, followed by normalization to the mean value of standardized Slo1 signals in WT mice. Data are shown as mean ± SEM. P < 0.05. (B) Western blot analysis for endogenous Slo1 levels in indicated cells transfected with siRNA for Ctrl or DDB1. Quantification of Slo1 was normalized to ß-tubulin. (C) Co-immunoprecipitation of Slo1 with WT or mutant CRBN and DDB1 in CRBN KO 293T cells. Cells were transfected with expression vectors encoding indicated genes followed by IP with Flag affinity beads. (D) CHX (100 µg/ml) chasing assays for the stability of Slo1 in WT or CRBN KO 293T cells. Cells were transfected with equivalent Flag-Slo1 along with equivalent GFP and harvested at the indicated time following CHX (Left).Arrowhead indicates the band of Slo1 used for quantification. Plot shows the quantification of Slo1 normalized to GFP signals and relative to the normalized Slo1 signals at the 0 min time point (Right). (E) Western blot analysis for Slo1 levels in LN229 glioma cells transfected with siRNA for ctrl or CRBN. Quantification of Slo1 was normalized to ß-actin. (F) Western blot analysis for Slo1 levels in U87mG and LN229 glioma cells after CRISPR/Cas9-mediated CRBN gene editing. Quantification of Slo1 was normalized to ß-actin. (G) Western blot analysis for Slo1 levels in cytoplasm or membrane extracts prepared from WT or CRBN KO U87mG cells. Pan-Cadherin, as a marker of membrane proteins. Quantification of membrane Slo1 was normalized to Pan-Cadnerin. (H) Western blot analysis for Slo1 levels in CRBN KO U87mG cells stably expressing vehicle and WT or mutant CRBN. Quantification of Slo1 was normalized to ß-actin.

Fig 5: DDB1 deletion in mouse brain causes learning and memory defects.All behavior analysis data are shown as mean ± SEM. (A) Novel object recognition to test texture-associated short-term memory in mice. Contacts of mice towards the familiar (A) or novel (C) object were recorded after training. Ddb1F/F;Camk2a-Cre mice could not differentiate object C from A while Ddb1F/F mice could (Ddb1F/F n = 11, Ddb1F/F;Camk2a-Cre n = 10, P = 0.026). (B) Barnes maze. The time for Ddb1F/F; Camk2a-Cre mice (n = 10) or Ddb1F/F mice (n = 11) to enter the escape tunnel were recorded during the training days (P < 0.05 for each block). Two-way repeated-measures ANOVA (genotype versus day) revealed a genotype effect (F(1,19) = 6.518, P = 0.0194) and day effect (F(2,38) = 6.318, P = 0.0043) for both groups. (C) Barnes maze. The errors (numbers of entrance into wrong tunnel) made by mice were recorded during the training days (P = 0.014 for block day 7–9). Two-way repeated-measures ANOVA revealed a genotype effect (F(1,17) = 13.68, P = 0.0018) for both groups. (D) Barnes maze. Incidence distribution of strategy for mice to search the escape tunnel were recorded during the training days. Two-way repeated-measures ANOVA revealed a significant differences in strategy use (F(3,16) = 29.00, P < 0.0001) and a genotype × strategy effect (F(3,16) = 13.16, P = 0.0001) for both groups. (E) Barnes maze. The time for mice spent in the target quadrant (where the removed escape tunnel was located) were recorded on the probe day (P = 0.0027).