Previously to detect CDKL activity
Previously, to detect CDKL5 activity, we used a radioactive phosphorylation assay with amphiphysin 1 protein as a substrate [,,]. However, it is necessary to establish a safe and rapid method for analyzing CDKL5 mutant activity. Here, we report a new method for investigating autophosphorylation activity of CDKL5 by phos-tag SDS–PAGE. The detailed method is shown in A. The phos-tag captures phosphorylated molecules and SDS–PAGE analysis separates the phosphorylated protein from the unphosphorylated species as a slower migrating band. In a previous report, full-length CDKL5 was solely expressed in the insoluble fraction . As observed in this previous report, when CDKL5(WT) was expressed in the recombinant protein was exclusively obtained in the insoluble fraction (B). Various protein kinases, such as casein kinase 1, are autophosphorylated in protein Isoprenaline HCl australia systems . Hence, we first determined whether CDKL5 undergoes activity-dependent autophosphorylation in , which is detectable by phos-tag SDS–PAGE. CDKL5(WT) or CDKL5(KD) were expressed in precipitated fractions and analyzed by phos-tag SDS–PAGE. A slower migrating band was only observed in the WT lane (C, arrow). Nonetheless, quantitative analysis showed that a phosphorylated protein signal with a relative intensity 10% that of the WT sample was observed in the CDKL5(KD) lane. CDKL5(KD) showed no activity in a previous study using a conventional method . Therefore, these results indicate that ~10% signal intensity in our method represents the background noise threshold. To confirm that this band shift was caused by phosphorylation, CDKL5(WT) expressed in precipitated fractions was treated with λPPase-His, a known serine/threonine/tyrosine phosphatase, for 1 h at 30 °C. The shifted band disappeared with this treatment in WT lane, while KD band pattern was not changed by λPPase-His (D). These results show that CDKL5 is autophosphorylated in , and that this phosphorylation is detectable by phos-tag SDS–PAGE. Next, we determined whether the method is applicable for analysis of CDKL5 mutants. We have already reported that CDKL5(Y177C) (identified from a Japanese patient with atypical Rett syndrome) completely loses its phosphorylation activity . Thus, CDKL5(Y177C) activity was analyzed using our new approach. The band shift was only detected in the WT lane, and not the Y177C lane (A). To confirm the effect of the CDKL5 missense mutation, we chose a single amino acid mutation in the kinase domain from RettBASE , a CDKL5 mutant database, and a recent study . In this study, ten pathogenic or likely pathogenic variants and four variants of uncertain significance were selected from within the catalytic domain (B). As shown in B, human and mouse catalytic domains are completely conserved, therefore we used a mouse CDKL5 expression plasmid instead of a human one. The mutants were analyzed by our method, and all variants showed dramatically reduced catalytic function (C). Only the H145Y mutant displayed partial kinase activity with a relative intensity of the shifted band of 36% compared to that of the WT protein. These results should be confirmed by an kinase assay using an exogenous substrate. Subsequently, we determined whether this loss of activity is dependent on an amino acid substitution. Accordingly, we analyzed benign or likely benign point mutants (H36R) using our method . Western blotting showed that the band shift was observed in both the WT and non-pathogenic mutant lanes (D), suggesting that correct autophosphorylation has occurred. To date, over 50 missense mutations that might be related to disease onset have been identified in the catalytic domain of CDKL5 . Nevertheless, only a few studies have examined their activity [, , ]. Although analysis tools for protein function are easy to use and provide good predictive value, direct demonstration of protein activity can overcome prediction consequences. However, previous activity detection methods rely on purification (e.g., of CDKL5 enzyme and substrates), and involve use of dangerous RI. The new method described here is an efficient and convenient tool for initial screening to determine whether or not a gene mutation is pathogenic. Furthermore, not only autophosphorylation activity but also measurement of phosphorylation activity as an exogenous substrate is necessary for precise enzymatic activity.