Tudor-domain protein PHF20L1 reads lysine methylated retinoblastoma tumour suppressor protein

Abstract

The retinoblastoma tumour suppressor protein (pRb) classically functions to regulate early cell cycle progression where it acts to enforce a number of checkpoints in response to cellular stress and DNA damage. Methylation at lysine (K) 810, which occurs within a critical CDK phosphorylation site and antagonises a CDK-dependent phosphorylation event at the neighbouring S807 residue, acts to hold pRb in the hypo-phosphorylated growth-suppressing state. This is mediated in part by the recruitment of the reader protein 53BP1 to di-methylated K810, which allows pRb activity to be effectively integrated with the DNA damage response. Here, we report the surprising observation that an additional methylation-dependent interaction occurs at K810, but rather than the di-methyl mark, it is selective for the mono-methyl K810 mark. Binding of the mono-methyl PHF20L1 reader to methylated pRb occurs on E2F target genes, where it acts to mediate an additional level of control by recruiting the MOF acetyltransferase complex to E2F target genes. Significantly, we find that the interplay between PHF20L1 and mono-methyl pRb is important for maintaining the integrity of a pRb-dependent G1-S-phase checkpoint. Our results highlight the distinct roles that methyl-lysine readers have in regulating the biological activity of pRb.

Figure 1

Identification of a new reader protein for pRb methylated at K810. (a) Schematic representation of pRb and PHF20L1 proteins. The amino acid sequence around residue K810 (in red) is expanded to indicate the CDK consensus motif SPLK (boxed). A methyl-dependent interaction with the tudor 1 domain of PHF20L1 is indicated. The amino acid sequence of PHF20L1 between residues 18 and 53 is displayed to highlight tudor 1 domain residues important for methyl-lysine recognition (*). The D23 and F47 residues mutated to alanine in this study are also indicated. (b) CADOR array probed with anti-GST (top), biotinylated RbK810me0 (middle) or biotinylated RbK810me1 (bottom). The grey boxed regions demarked show binding of the methylated pRb peptide to the tudor 1 domain of PHF20L1. The additional green spots represent the previously described interaction with 53BP1. (c) Peptide-binding assay in which RbK810me0 or RbK810me1 peptide was incubated with recombinant GST-PHF20L1 tudor 1. The left hand side displays flow-through from the assay, while the right hand side displays the remaining eluted protein. n=3. (d) As above, although RbK810me0, -me1, -me2 and -me3 peptides were used. n=2. (e) Biolayer interferometry real-time kinetic analysis of immobilised RbK810me0, -me1, -me2 and -me3 peptides bound to His-PHF20L1 tudor 1. (f) As above, but showing the concentration dependent binding of PHF20L1 tudor 1 with the RbK810me1 peptide. A K_D value of 28 μM was calculated from these data. (g) Peptide-binding assay in which RbK810me0 or RbK810me1 peptides were incubated with recombinant GST-PHF20 (tudor 1, tudor 2 or tudor 1+2) or GST-PHF20L1 (tudor 1, tudor 2 or tudor 1+2). n=3

Figure 2

The tudor 1 domain of PHF20L1 shows specificity for RbK810me1. (a) SAOS2 cells were transfected with 3 μg of HA-pRb/HA-pRb-K810R and 1 μg of Flag-PHF20L1/empty vector as indicated. Cells were also treated with 20 μM etoposide for 16 h where appropriate. An immunoprecipitation was performed using anti-Flag agarose and co-precipitating pRb was detected by immunoblot. The numbers below the blot indicate the relative amount of pRb co-immunoprecipitated with PHF20L1. n=3. (b) Extracts from U2OS pRb CRISPR cell lines stably transfected with ectopic empty vector (−), HA-pRb or HA-pRb-K810R were used in an immunoprecipitation with anti-HA agarose. Co-precipitating endogenous PHF20L1 was analysed by immunoblot. The numbers below the blot indicate the relative amount of PHF20L1 co-immunoprecipitated with pRb. n=2. (c) SAOS2 cells were transfected with 3 μg of HA-pRb and 1 μg of Flag-PHF20L1, Flag-PHF20L1 D23A or Flag-PHF20L1 F47A as indicated. An immunoprecipitation was performed with anti-Flag agarose and co-immunoprecipitating pRb was detected by immunoblot. (d) Peptide-binding assay in which RbK810me0 or RbK810me1 peptides were incubated with recombinant GST-PHF20 tudor 1, GST-PHF20L1 tudor 1 or GST-PHF20L1 tudor 1 D23A. The left hand side displays flow-through from the assay, while the right hand side displays the remaining eluted protein. n=3

Figure 3

PHF20L1 recruits MOF to methylated pRb. (a) U2OS cells were transfected with 3 μg of HA-pRb/HA-pRb-K810R and 1 μg of Flag-MOF/empty vector as indicated. Cells were treated with 20 μM etoposide for 16 h where appropriate. An immunoprecipitation was performed using anti-Flag agarose and co-precipitating pRb was detected by immunoblot. The numbers below the blot indicate the relative amount of pRb co-immunoprecipitated with MOF. n=2. (b) In vitro interaction assay in which 250 ng of His-PHF20L1 was incubated with 250 ng of GST-MOF or GST. His-PHF20L1 was immobilised on Ni-NTA agarose and co-precipitating MOF was detected by immunoblot. n=3. (c) Peptide-binding assay in which RbK810me1 peptide was incubated with GST-MOF and His-PHF20L1 as indicated. The left hand side displays flow-through from the assay, while the right hand side displays eluted protein. n=3. (d) U2OS cells were transfected with 20 nM control or PHF20L1 siRNA, followed by 3 μg HA-pRb and 1 μg Flag-MOF/empty vector as indicated. An immunoprecipitation was performed using anti-Flag agarose and co-precipitating pRb was detected by immunoblot. The numbers below the blot indicate the relative amount of pRb co-immunoprecipitated with MOF. n=4. (e) U20S cells were transfected with 2 μg empty vector (−), Flag-PHF20L1 or Flag-MOF. An immunoprecipitation was performed with anti-Flag antibody and chromatin was analysed by PCR using primers targeting the indicated promoters. n=3. (f) U2OS cells were transfected with 2 μg of Flag-PHF20L1 and Flag-MOF. Extracts were immunoprecipitated with control IgG or PHF20L1 antibodies (first ChIP). The immunoprecipitated chromatin was then re-immunoprecipitated a second time with control IgG or MOF antibodies, as indicated (second ChIP)

Figure 4

pRb-dependent recruitment of PHF20L1-MOF to the chromatin of E2F-responsive promoters by ChIP analysis. (a) U2OS cells were transfected with 20 nM control or pRb siRNA as indicated. Cell extracts were then immunoprecipitated with control IgG, PHF20L1, pRb or H4K16ac antibodies, and chromatin was analysed by qPCR using primers targeting the TS and DHFR promoters. n=3 (b) U2OS cells were transfected with 20 nM control or MOF siRNA as indicated. Cell extracts were then immunoprecipitated with control IgG or H4K16ac antibodies, and chromatin was analysed by qPCR using primers targeting the indicated promoters. n=2. (c) U2OS or U2OS pRb CRISPR cells were prepared for ChIP analysis, and immunoprecipitated with control IgG, PHF20L1 or H4K16ac antibodies. Chromatin was analysed by qPCR using primers targeting the indicated promoters. n=2. (d) U2OS cells were transfected with 20 nM control or PHF20L1 siRNA as indicated. Chromatin immunoprecipitation was then performed using primers against (i) CDC25A, (ii) TS or (iii) DHFR promoters. n=3

Figure 5

Characterisation of PHF20L1 in cells. (a) U2OS cells were seeded on coverslips and transfected with 1 μg of Flag-PHF20L1. Cells were also treated with 20 μM etoposide or 1 mM hydroxyurea for 24 h where indicated. In some cases, cells were released from hydroxyurea block for the indicated time points. Cells were fixed and prepared for immunofluorescence. A flow cytometry analysis of cells is included in Supplementary Figure S2a to demonstrate cell synchronisation. (b) U2OS cells were transfected with 20 nM control siRNA (C), or siRNA-targeting PHF20L1 (P). Cells were prepared for flow cytometry analysis. An immunoblot was also performed to monitor input protein levels. n=5. (c) As above, though cells were treated for 24 h with 2 μM PD0332991 (Cdk4/Cdk6 inhibitor). n=3

Figure 6

Impact of PHF20L1 and pRb on cell cycle progression. (a) (i) U2OS cells were transfected with 20 nM control siRNA (C), or siRNA-targeting PHF20L1 (P), pRb (R) or a combination of both (PR). Some cells were treated 24 h with 1 mM hydroxyurea and released for 2 h. Cells were labelled with BrdU and prepared for flow cytometry analysis. (ii) The total percentage change of BrdU-positive cells between the hydroxyurea treated and released samples was calculated and displayed. (iii) An immunoblot was performed to monitor input protein levels. n=2. (b) (i) Representative cell cycle profiles taken from a flow cytometry analysis experiment in which U2OS or U2OS pRb CRISPR cell lines (Rbcr) were transfected with 20 nM control siRNA (C) or siRNA-targeting PHF20L1 (P). Cells were treated 42 h with 1 mM hydroxyurea and some released for 8 h. Numbers indicate the mean percentage of cells in S+G2/M phases from the technical repeats within this representative experiment, with S.D. shown. Student’s t-tests performed from independent biological replicates indicated that the difference between the U2OS control siRNA sample and the U2OS PHF20L1 siRNA, pRb CRISPR control siRNA, or pRb CRISPR PHF20L1 siRNA samples were all statistically significant (P<0.02). (ii) The total percentage change of cells in S+G2/M phases between hydroxyurea treated and released samples was calculated and displayed. (iii) An immunoblot was performed to monitor input protein levels. n=3. (c) Model for PHF20L1-MOF assembly with methylated pRb on chromatin. In response to pRb mono-methylation, PHF20L1 is recruited to chromatin-bound pRb, where it acts to regulate a pRb-dependent G1–S-phase checkpoint. This checkpoint likely involves the acetyltransferase activity of co-recruited MOF complex, which can target H4K16 at the promoters of E2F-responsive genes (i). In the absence of PHF20L1 (ii) or pRb (iii), this checkpoint response is lost, and cells enter S phase in an inappropriate manner