Role of Zbtb24 and DNA methylation in late B-cell differention and humoral memory – METHYL-MEMORY

Several models are being established simultaneously: mice that bear either knock-in mutations found in ICF1 (Dnmt3b) or ICF2 (Zbtb24) patients, or conditional knockout alleles for Zbtb24 or Lsh (ICF4). In parallel, a retroviral-based RNA interference approach is developped against Zbtb24. These models will be explored for their developmental and immune capacities.
2) Genome-wide methylome analysis
Methylation profile of blood cells form the different categories of ICF patients is determined with Infinium Human Methylation 450 Bead Chip Kit. In mice, this profile is recorded with 24k oligo v1.0 microarrays.
3) Zbtb24 interactors
A yeast two-hybrid screen is conducted and candidate interactors will be validated by DuoLink and co-immunoprecipitation

1) Zbtb24 is vital for mouse early embryonic development
Zbtb24 loss-of-function is lethal in mouse, at an early embryonic developmental stage ; this result was not anticipated, given the phenotype of ICF2 patients. Unfortunately, the conditional knockout model is not valid as no homozygous floxed embryo could ever be recovered.
2) ICF mutations and DNA methylation
An ES cell line was derived from homozygous Zbtb24 knock-in mutant blastocysts ; its analysis confirmed the role of Zbtb24 in the maintenance of DNA methylation at centromeric and pericentromeric satellite repeats. Genome-wide methylome analysis of peripheral blood cells from ICF patients confirmed that Zbtb24 is required for the maintenance of DNA methylation profile in human somatic cells as well as for the maintenance of parental imprinting. Moreover, the identity of hypomethylated genes indicated that ICF2, 3 and 4 patients cluster apart from ICF1 patients, which suggest that Zbtb24, Lsh and CdcA7 play also a role in a common pathway that does not depend upon Dnmt3b. The yeast two-hybrid screen is almost completed and putative interactors will be validated thereafter.

Alternative models of ICF syndrome are currently being developped to overcome the problem of embryonic lethality caused by Zbtb24 loss-of-function : a new Zbtb24 cKO mouse strain, a Lsh cKO mouse strain (in collaboration with K. Muegge's laboratory, USA) and an RNAi strategy against Zbtb24. These models will be used to assess which late B-cell differentiation steps are affected in ICF syndrome.
Hypo- and hyper-methylated targets recorded through our genome-wide studies will be validated and compared to the dataset generated from the Zbtb24 mutant ES cell line. The impact of the deregulation of such targets on the humoral immune response will be eventually tested in mouse.
Putative Zbtb24 physical interactors will be validated in human cells in vitro and the functional consequences of these interactions on DNA methylation and heterochromatinization will be explored.

in progress

Memory B cells (Mem B) and long-lived plasma cells (LLPC) ensure the long-term antibody-mediated protection against pathogens. However, the factors that support their differentiation and sustain their exceptional longevity remain largely unknown.
We will address this question through the study of the human ICF syndrome (Immunodeficiency, Centromeric instability and Facial abnormalities), a rare genetic disease characterized by a severe decrease in both immunoglobulin titers and Mem B cells numbers. ICF is caused by mutations in DNMT3B (ICF1), a de novo DNA methyltransferase, or ZBTB24 (ICF2), a putative transcription factor. ZBTB24, together with DNMT3B, could play a role in DNA methylation since ICF patients display hypomethylation of juxtacentromeric satellite repeats which causes centromeric instability. As the ICF phenotype is consistent with an alteration of LLPC and Mem B cell compartments, elucidating the link between its genetic basis and the immunodeficiency should be very informative.
We will first determine which stage of B cell differentiation is affected by ICF mutations. We will study their impact on in vitro differentiation of human B cells into PC and/or Mem B cells, using both cells from patients and RNA interference. In parallel, two new Zbtb24 mutant mouse strains will be characterized: one with a gene-trapped allele that can be converted into a conditional knockout or a reporter (LacZ) allele, and one with knock-in mutations that reproduce those of an ICF2 patient. We will confirm that both recapitulate the ICF phenotype and Zbtb24 expression will be tracked with the LacZ allele. The lymphoid development of Zbtb24 mutants and their ability to mount memory antibody responses will be analyzed and we will assess if this defect is B-cell intrinsic. At last, we will determine if Dnmt3b deficiency similarly impairs late B-cell differentiation, by lymphoid reconstitution with fetal liver from our mouse strain with hypomorphic mutations found in ICF1 patients.
Our second aim is to delineate how Zbtb24 deficiency affects B cells. To identify Zbtb24 direct targets, we will perform transcriptome and methylome analyses on the mutant mice and compare these datasets to those generated earlier from the Dnmt3b deficient mice described above. In addition, Zbtb24 deficiency could alter the heterochromatinization of juxtacentromeric regions. This defect, combined with the spontaneous repression of p53 in germinal center B cells, could lead to the transcription of satellite-derived dsRNA in this cell type and the subsequent triggering of a type-I interferon mediated apoptotic pathway. We will test this hypothesis by breeding our mice with a mouse strain deficient for the type-I IFN receptor. Alternatively, hypomethylation and transcription of satellite repeats could lead to perturbed nuclear organization, and ultimately, to perturbed expression programs associated with B-cell identity and fate. We will characterize transcription of satellite repeats and nuclear organization, especially heterochromatin compartments, in the nuclei of progenitors and B cells from mouse models.
Our third aim is to search for Zbtb24 binding partners to understand how it regulates DNA methylation. We will undertake a global yeast double-hybrid screening and we will validate in mammalian cells the interaction with several candidate proteins with known implication in DNA methylation and assembly of heterochromatin. Using loss of function approaches, we will then define whether these factors are implicated in the recruitment of ZBTB24 or DNMT3B or both at heterochromatin regions and target genes identified in Aim 2.
In conclusion, our study should refine our knowledge on B-cell differentiation and pathways to DNA methylation. The long-term outputs could be the improvement of the humoral responses of ICF patients, and more generally a better understanding of B-cell memory formation, a key knowledge for the development of efficient vaccines.