Fig 2: Co-localisation of MKS1 and UBE2E1 under conditions of ciliary resorption.(a) MKS1 (green) and UBE2E1 (red) partially colocalise at the basal body/centrosome in human wild-type hTERT-RPE1 cells, particularly when induced to resorb cilia by treatment with 10% FCS after 72 hr of serum starvation. White arrowheads indicate cells magnified in insets. Scale bar = 10 µm. (b) Bar graph indicates the percentage of cells in which MKS1 and UBE2E1 co-localise at the basal body (black), and the percentage without co-localisation (grey) for three independent biological replicates, with examples shown of representative cells. (c) Figure details as for (a) showing partial co-localisation of MKS1 and UBE2E1 in human ARPE19 cells. (d) Bar graph details as for (b). Data in (b) and (d) were analysed by two-way ANOVA followed by Tukey’s multiple comparison test (statistical significance of comparisons indicated by ** p < 0.01, *** p < 0.001).

Fig 3: MKS1 and UBE2E1 interact to regulate ß-catenin ubiquitination.(a) Immunoblot showing increased co-immunoprecipitation of ß-catenin by anti-MKS1 in shUbe2e1 knockdown cells compared to shScr control cells. (b) TUBE pulldown followed by ß-catenin immunoblotting, showing polyubiquitination of ß-catenin was increased following MKS1 knockdown and in the presence of the wild-type (WT) UBE2E1. Inactive form of UBE2E1 (DN) had not had effect on polyubiquitination of ß-catenin, highlighting the importance of this UBE2E1 in ß-catenin degradation. (c) Immunofluorescence staining of hTERT-RPE1 cells showing co-localisation of phosphorylated (P)-ß-catenin (green) with ?-tubulin (red) at the base of cilia (arrowheads). P-ß-catenin localisation significantly increased following siMKS1 knockdown (paired two-tailed Student t-test, **** p < 0.0001 for three independent biological replicates; > 40 cells quantified per replicate). Scale bar = 5 µm. (d) Schematic representation of UPS regulation of MKS1 and ß-catenin protein levels at the ciliary apparatus. Protein levels of MKS1 (pink) and UBE2E1 (light brown) are co-dependant through regulation at the base of the cilium. MKS1 localises to the TZ (dashed pink lines) and is mono-/bi-ubiquitinated by a complex that includes UBE2E1 and RNF34 (blue). MKS1 and UBE2E1 regulate each other, what has an effect on downstream UBE2E1 role in regulation of polyubiquitination of ß-catenin (yellow). The correct regulation between these proteins facilitates normal proteasomal function and canonical Wnt signalling (small pink arrow). Both processes are de-regulated following MKS1 mutation of loss (red cross), causing aberrant accumulation of UBE2E1 and polyubiquitinated ß-catenin and disrupted tethering to the ciliary apparatus. Figure 7—source data 1.MKS1 and UBE2E1 interact to regulate ß-catenin ubiquitination: full western blots.

Fig 4: The E2 ubiquitin conjugation enzyme UBE2E1 interacts with MKS1.(a) Domain structure of MKS1 and UBE2E1 proteins for the indicated isoform showing the locations of the B9/C2 domain, putative ubiquitinated lysines in blue (predicted by UbPred), a predicted coiled-coil (CC) motif, and the E2 ubiquitin (UBQ) conjugation domain in UBE2E1. Numbering indicates the amino acid residue. Dashed lines indicate the region used as ‘bait’ in MKS1 for the yeast two-hybrid assay and the ‘prey’ clones in the UBE2E1 interactant. (b) List of preys identified in the MKS1 Y2H screen (c) Left panel: yeast ‘one-to-one’ assays for the indicated bait, prey and control constructs. Right panel: only colonies for the positive control (p53+ SV40 large T) and MKS1 bait+ UBE2E1 prey grew on triple dropout (-Leu -Trp -His) medium. (d) GST-UBE2E1 purified from bacterial extracts (left panel) pulled down endogenous MKS1 from ARPE19 whole cell extract. (e) Co-immunoprecipitation (co-IP) of endogenous UBE2E1 by rabbit polyclonal anti-MKS1, but not pre-immune serum or an irrelevant antibody (Ab; anti cmyc); IgG light chain (LC) is indicated. (f) Co-IP of exogenously expressed cmyc-MKS1 by anti-UBE2E1 but not pre-immune serum or an irrelevant antibody. Figure 2—source data 1.The E2 ubiquitin conjugation enzyme UBE2E1 interacts with MKS1: full western blots.

Fig 5: Co-dependant regulation of MKS1 and UBE2E1.(a) SUPER-TOPFlash assays in wild-type or MKS1-mutated fibroblasts, following transient co-transfection with either exogenous control, MKS1-cmyc or UBE2E1-FLAG-cmyc, and treatment with either Wnt3a or control conditioned medium. Statistical significance of the indicated pairwise comparisons with control for three independent biological replicates are shown (* p < 0.05, ** p < 0.01, *** p < 0.001, **** p < 0.0001; unpaired two-tailed Student t-test; error bars indicate s.e.m.) (b) SUPER-TOPFlash assays in shScr and shUbe2e1 cell-lines, following transient co-transfection with either exogenous cmyc-MKS1 or empty plasmid construct (control) and treatment with either Wnt3a or control conditioned medium, as indicated. Statistical comparisons as for (a). (c) Top panel: increased per cell staining intensity for UBE2E1 following MKS1 siRNA knockdown Bottom panel: increased per cell staining intensity for MKS1 in Ube2e1 mIMCD3 knockdown cells Scale bars = 10 µm. Bar graphs quantitate staining intensities for three independent biological replicates. Statistical significance of pairwise comparisons as for (a), error bars indicate s.e.m. Western blots (panels on right) show increased UBE2E1 protein levels for siMKS1 knockdown cells, and increased MKS1 protein levels for shUbe2e1 cells. Quantitation of band intensities were normalised to ß-actin loading control. (d) HEK293 cells were transiently transfected with control vector (-), constant (+) or high (+++) levels of cmyc-MKS1 and/or FLAG-cmyc-UBE2E1. Levels were normalised to ß-actin loading control. MKS1 levels moderately decreased with increasing levels of UBE2E1, whereas high levels of MKS1 caused loss of UBE2E1. Figure 5—source data 1.Co-dependant regulation of MKS1 and UBE2E1: full western blots.