CDKL5 kinase controls transcription‐coupled responses to DNA damage

Abstract Mutations in the gene encoding the CDKL5 kinase are among the most common genetic causes of childhood epilepsy and can also give rise to the severe neurodevelopmental condition CDD (CDKL5 deficiency disorder). Despite its importance for human health, the phosphorylation targets and cellular roles of CDKL5 are poorly understood, especially in the cell nucleus. Here, we report that CDKL5 is recruited to sites of DNA damage in actively transcribed regions of the nucleus. A quantitative phosphoproteomic screen for nuclear CDKL5 substrates reveals a network of transcriptional regulators including Elongin A (ELOA), phosphorylated on a specific CDKL5 consensus motif. Recruitment of CDKL5 and ELOA to damaged DNA, and subsequent phosphorylation of ELOA, requires both active transcription and the synthesis of poly(ADP‐ribose) (PAR), to which CDKL5 can bind. Critically, CDKL5 kinase activity is essential for the transcriptional silencing of genes induced by DNA double‐strand breaks. Thus, CDKL5 is a DNA damage‐sensing, PAR‐controlled transcriptional modulator, a finding with implications for understanding the molecular basis of CDKL5‐related diseases.

▸ Figure EV2. Restricting CDKL5 expression to the cell nucleus.
A Extracts of CDKL5-disrupted U-2-OS (Flp-In T-REx) cells (CDKL5 D/D ) stably expressing CDKL5 NLS WT or a K 42 R kinase-dead mutant (CDKL5 NLS -KD) or empty vector were subjected to Western blotting with the antibodies indicated. Two different dishes of cells are shown per condition. B Subcellular fractionation of lysates from CDKL5 D/D cells stably expressing CDKL5, CDKL5 NLS WT or CDKL5 NLS KD or empty vector. Lysates were fractionated to isolate proteins found in the following subcellular compartments: cytoplasmic (Cyt), membrane (Mb), nuclear (Nuc), chromatin (Ch) or cytoskeleton (Csk). Fractionated samples were resolved by SDS-PAGE and probed with antibodies shown. C CDKL5 D/D cells stably expressing CDKL5, CDKL5 NLS WT or CDKL5 NLS KD were subjected to indirect immunofluorescence analysis with anti-CDKL5 antibodies. Scale bar is 10 lm. D CDKL5 D/D cells stably expressing CDKL5 NLS WT or CDKL5 NLS KD (or empty vector) were treated with 500 µM H 2 O 2 for 15 min. Samples were resolved by SDS-PAGE and probed with indicated antibodies or stained with Ponceau S to show equal loading. Rep=biological replicate. E Peptide kinase assays to investigate CDKL5 sequence specificity. Anti-FLAG precipitates from HEK293 cells transiently expressing FLAG-tagged CDKL5 (wild-type "WT" or a K 42 R kinase-dead "KD" mutant) were incubated with synthetic peptides corresponding to sequence around the previously reported CDKL5 phosphorylation site in MAP1S (Ser 900 ) designed specifically to investigate the effect of amino acid substitutions A 901 on the phosphorylation of MAP1S Ser 900 . Assays were done in the presence of [c-32 P]-labelled ATP-Mg 2+ , and peptide phosphorylation was measured by Cerenkov counting. Phosphorylation of the control wild-type MAP1S peptide is taken as 100% (*). The data are represented as mean AE SEM from three independent experiments. The RPXSA motif is shaded in blue, and amino acid substitutions compared with the wild-type MAP1S Ser 900 peptide are shown in red.
Source data are available online for this figure.   Figure EV3. Phosphoproteomic data quality control.
A Left: Normal Q-Q plot of the raw TMT intensity data with large deviations from a normal distribution as seen from datapoints not following the indicated line in the plot. Right: Q-Q plot of the TMT intensity data after VSN transformation. Only minor deviations from the line indicates the transformed data follow a normal distribution to a satisfactory degree. The hypervariable datapoints in the upper quantiles are controlled by the application of the robust implementation of the empirical Bayes algorithm used by limma (Phipson et al, 2016) and implemented in the analysis scripts. B Standard deviation plotted against the intensity rank of the VSN-transformed TMT data. Red line indicates the mean standard deviation. Line is approximately horizontal, indicating that the variance is not overly dependent on intensity rank and suggests a successful VSN transform. C Boxplot of intensity distribution in each TMT channel. No obvious discrepancy between the median values of the individual channels indicates a successful calibration by VSN and no introduction of an obvious intensity bias for any experimental group. The central band of the boxplot indicates the median value, while the hinges represent the first and third quartile (bottom and top of boxplot, respectively). The whiskers extend to the largest/smallest (upper or lower whisker, respectively) datapoint not further than 1.5 times the interquartile range from their respective hinge. The experiment was conducted using five biological replicates of CDKL5 NLS WT (WT, red) and CDKL5 NLS KD (KD, blue) where each TMT channel represents a single biological replicate from the respective group.
Source data are available online for this figure.
Taran BrdU-sensitized U-2-OS (Flp-In T-REx) cells were subjected to nuclear line micro-irradiation (355 nm). Cells were fixed and then mock-treated (con) or treated with lambda-phosphatase prior to incubation with the primary antibodies indicated. Alternatively, ELOA-pSer 311 phosphopeptide was included during incubation with the primary antibodies before indirect immunofluorescence analysis. Quantification of ELOA-pSer 311 signal at the laser tracks is shown. Data represent mean AE SD of total pELOA Ser 311 intensities in different biological replicates as indicated (n). For simplicity, only intensities greater than zero are shown. Statistical significance was assessed by one-way ANOVA test. Asterisks **** indicate P-values of < 0.0001. Scale bar is 10 lm.   A CDKL5, ATM and ZMYND8 were depleted in U-2-OS 263 IFII reporter cells using indicated siRNA. siCON-non-targeting control. B qRT-PCR measurements in U2OS I-PpoI cells to validate the silencing efficiency of siCDKL5 and the subsequent rescue efficiency of ectopic expression of siRNAresistant forms of CDKL5 wild-type (WT) and K 42 R kinase-dead mutant (KD). P2 and P3 primer pairs were used to analyse the changes in mRNA levels. The mean AE SD from two qPCR replicates of two independent experiments is shown.
Source data are available online for this figure.