Background: The eucaryotic genome is organized in chromatin domains which affect its function. This organization is partly established by special architectural and transcriptional factors specific to the cellular context. In an hematopoietic context, PLZF (Promyelocytic Leukemia Zinc Finger protein), a member of the family of POK repressor proteins, is directly implicated in the regulation of epigenetic modifications by tethering DNA methyltransferases (DNMT) and histone deacetylases (HDAC) to specific genomic targets. We have previously shown that PLZF transcriptional activation is controlled by acetylation of specific lysine residues (K647/650/653) affecting its nuclear localization. PLZF is mainly expressed in CD34 positive cells and has been shown to be crucial in the maintenance of hematopoietic stem cells (HSC). However, the epigenetic role of PLZF in HSC protection and maintenance is not yet understood.

Methodology: We created PLZF functional knock-in mouse models by introducing PLZF lysine mutants with altered epigenetic functions at the PLZF locus. We used purified bone marrow (BM) cells from 12 week old mice to assess their self-renewal capacity by methylcellulose-based medium serial replating assays to detect and quantify hematopoietic progenitors in colony-forming unit (CFU) experiments and long-term self-renewal. In parallel, Methyl DNA immunoprecipitation followed by deep-sequencing (MeDiP-seq) was performed in order to establish DNA methylation pattern and characterize PLZF induced DNA methylation regions (PLZF-DMR). Retrotransposon activity was determined by quantifying the expression of L1 retrotransposon mRNAs in mouse cells and by retrotransposon assay using a GFP-L1 reporter in human 293T cells. Validation and functions of PLZF genomic targets found in mouse cells were tested in KG1a CD34+ hematopoietic human cells by Chromatin immunoprecipitation (ChIP) and reporter gene assays using PLZF-DMR luciferase vectors containing L1 and also Alu and telomeric sequences targeted by PLZF.

Results: We have created two PLZF functional mutants, PLZFON mutant constitutively binding DNA and PLZFOFF mutant unable to interact with DNA, and shown that PLZFOFF mice are infertile due to loss of germinal cells recapitulating the PLZF knock-out mouse model phenotype. In both PLZFOFF and PLZFON 12 week old mice the number of myeloid colonies decreased by 62% compared to PLZFWT and a total absence of self-renewal capacity at the 1st replating in PLZFOFF was noted. In BM cells, establishment of differential DNA methylation profiles by MEDIP-seq of PLZFOFF, PLZFON and PLZFWT mice allowed to identify more than 500 PLZF-DMRs. We found that primary PLZF genomic targets are repeat elements scattered throughout the genome (75% of all PLZF-DMRs). These repeat elements include retrotransposons (L1, 21.9%), Alu sequences (14.4%), retroviruses (22.1%) and sub-telomeric/telomeric regions (10%). We first investigated the impact of PLZF on L1 elements because of their involvment in genome instability, gene control and cancer through retrotransposition and methylation events: -a) L1 and PLZF interact through an 8bp conserved binding site; -b) PLZF binds to full length and truncated L1 DNA sequences, inducing DNA methylation and histone deacetylation propagation by recruiting of DNMT and HDAC proteins; L1 mRNA expression is increased PLZFOFF BM cells. Furthermore, in a retrotransposition assay using a GFP-L1 reporter in human cells we demonstrate that PLZF inhibits L1 retrotransposition and that PLZF-DMRs containing L1, Alu and telomeric sequences induce transcriptional repression by creating heterochromatin boundaries.

Conclusions: These results, using unique mouse models of PLZF inactivation, with loss of HSC maintenance show that PLZF may be a guardian of genome integrity of long living hematopoietic cells by establishing DNA methylation patterning, associated with constitutive heterochromatin domains involved in retrotransposition silencing and centromere/telomere stability.

Disclosures

No relevant conflicts of interest to declare.

Author notes

*

Asterisk with author names denotes non-ASH members.

Sign in via your Institution