Fig 2: Thermal stability assay. HEK293 cells stably expressing wild-type (wt) DYRK1A or L295F mutant were harvested and resuspended with PBS. Multiple aliquots of the cell suspensions were subjected to heat treatment for 3 min as indicated. Cells were lysed using Triton X-100 and cleared by centrifugation. The presence of the target protein in the soluble fraction was detected using western blot analysis and quantified. (A) Blots from a representative experiment. (B) Quantitative evaluation of four experiments (means and s.d.). For each pair of data points, differences in the amount of soluble protein were tested for statistical significance by one-sample t-test (*P<0.05).

Fig 3: Tyrosine autophosphorylation of wild-type and variant DYRK1A constructs. Wild-type (wt) and mutant GFP-DYRK1A fusion proteins were immunoprecipitated with a GFP antibody from cell extracts of transiently transfected HEK293 cells and subjected to immunoblot analysis. Tyrosine phosphorylation of the activation loop (Y321) was analyzed using an antibody against the activation loop tyrosine in HIPK2 that cross-reacts with the corresponding residue in DYRK1A (Abu Jhaisha et al., 2017). Relative tyrosine phosphorylation was calculated ratio of the pY321 signal relative to total amounts DYRK1A constructs as detected with a DYRK1A specific antibody. (A-C) Representative western blots from selected experiments, including overexposed images to demonstrate the absence of pY321 in the weakly expressed variants. (D) Quantitative evaluation of the catalytically active missense variants (means and s.d.; n=5 for wt versus L295F and n=6 for wt versus T588N). Differences were not statistically significant by one-sample t-test (P>0.05).

Fig 4: Effects of mutations on DYRK1A kinase activity. Wild-type (wt) and mutant GFP-DYRK1A fusion proteins were immunoprecipitated with a GFP antibody from cell extracts of transiently transfected HEK293 cells and subjected to radiometric immunocomplex kinase assays with the peptide substrate, DYRKtide. In vitro kinase activities were normalized to the amount of the respective GFP-DYRK1A protein as quantified by immunoblot analysis. Results are shown relative to wt DYRK1A (means and s.d.; n=5 for T588N, n=4 for L295F and D287N and n=3 for Y321F and K188R). Pairwise differences between wt and variants were tested for statistical significance by one-sample t-test (*P<0.05).

Fig 5: Reduced tyrosine autophosphorylation of in vitro-translated DYRK1A-L295F. DYRK1A constructs (28-499) were expressed in a cell-free E. coli-derived expression system. Coupled in vitro transcription-translation reactions were incubated for 90 min at 37°C before reaction products were subjected to immunoblot analysis. For quantification, pTyr signals were normalized to the total amounts of recombinant DYRK1A as determined by detection of the N-terminal StrepTag (means and s.d., n=3). The difference in relative tyrosine autophosphorylation was tested for statistical significance by one-sample t-test (*P<0.05).