Physical and Functional Association of PLZF with Histone Methyl Transferase and Histone Demethylase.

The promyelocytic leukemia zinc finger protein (PLZF) is a transcription factor fused to RARα in the t(11;17) translocation associated with retinoic acid resistant acute promyelocytic leukemia (APL). As a result of this chromosomal abnormality, two oncogenic proteins are produced, PLZF-RARα and RARα-PLZF. Wild type PLZF is expressed in CD34+ progenitor cells and declines during differentiation. PLZF is a tumor suppressor that causes cell cycle arrest, downregulating genes such as cyclinA2 and c-myc. We previously showed that transcriptional repression by PLZF is mediated by the recruitment of histone deacetylases to target genes, this being critical for its ability to control growth and affect RAR target genes. We now show that PLZF alters the methylation state of histones in its target genes. A biotinylated form of PLZF co-purified in cells along with a histone methyl transferase (HMT) activity for native histones. Using mutant histone H3 tail peptides, we showed that this activity methylated histone H3 on lysine 9 (H3K9me). Tagged forms of PLZF as well as endogenous PLZF co-precipitated in vivo with G9a histone methyl transferase, an enzyme that can mono and dimethylate H3K9 in euchromatin subject to gene repression. The interaction of PLZF with G9a required the presence of the N-terminal BTB/POZ domain as well as a second, more C-terminal, repression domain of PLZF. Given the newly found role of active histone demethylation in gene control we also tested the interaction of PLZF with LSD1, an enzyme associated with gene repression that demethylates H3K4. As in the case of G9a, the interaction of PLZF with LSD1 required both repression domains, suggesting, that these proteins may be part of a multi-protein complex containing multiple contact points with PLZF. Expression of G9a or LSD1 augmented transcriptional repression mediated by PLZF on reporter genes, indicating a functional interaction between histone methylation modifiers and PLZF. To determine the ability of PLZF to affect chromatin methylation in vivo, a Gal4-PLZF fusion protein was expressed in cells containing a chromatin-embedded Gal4-tk-Luciferase reporter gene. In the presence of PLZF, a chromatin immunoprecipitation experiment showed an increase in H3K9 methylation of the target gene while H3K4 methylation decreased, consistent with the ability of PLZF to interact with LSD1 and G9a. Lastly we compared the ability of the histone modifying proteins to interact with the APL fusion proteins PLZF-RARα, PML-RARα and NPM-RARα. Co-precipitation experiments showed a robust interaction between PLZF-RARα and G9a and LSD1 while the PML-RARα and NPM-RARα fusions bound these proteins significantly less avidly. Collectively all these data indicate that specific histone methylation is an important mode of action of PLZF in gene repression. The retinoic acid resistance of t(11;17)-APL may be related to its ability to interact with HMTs and histone demethylases. Hence therapeutic targeting of HMTs and histone demethylases might be considered as a novel mode of therapy in APL and other hematological malignancies.