One of the primary goals of our study is to understand the roles played by the transcription factor, IRF8 in hematopoietic cells. In the mouse, studies have focused on dendritic cells, B cells and macrophages, which are complemented by analyses of human B cell lines and normal B cells. We have studied expression of IRF8 during B cell development in the bone marrow (BM) from hematopoietic stem cells (HSC) to immature B cells. These studies demonstrated that the gene is expressed at low levels in HSC but at greatly increased levels in common lymphoid progenitors and the earliest cells committed to the B cell pathway (pre-pro-B), and then at reduced levels in more mature cells. IRF8 was shown to regulate expression of genes encoding PU.1 and EBF1, thereby promoting B lineage choice by HSC. This places IRF8 in the transcriptional regulatory network that includes PU.1, IKAROS, E2A, EBF and PAX5 and which is responsible for B cell lineage specification, commitment and differentiation. Autoreactive immature B cells in the BM are prevented from reaching peripheral lymphoid tissues by undergoing negative selection through deletion or by receptor editing to generate B cell receptors (BCR) that do not recognize self. They then move to the spleen as transitional B cells before being directed into the follicular (FOL) or marginal zone (MZ) B cell compartments. Studies of transitional B cells showed that they were no longer able to edit their receptors but were nonetheless still susceptible to signals governing positive and negative selection. We showed previously that in peripheral B cells, expression of IRF8 increases to highest levels in FOL B cell-derived germinal center (GC) B cells and then is strikingly downregulated in plasma cells. IRF8 contributes to the transcriptional regulation of two genes critically involved in GC B cell differentiation and function, BCL6 and AID. AID is the essential regulator of somatic hypermutation (SHM) and class witch recombination (CSR). Studies of AID-expressing cell lines with GC B cell features demonstrated that Igk V region genes were inducibly mutated at high rates, with a focus on A:T mutations that occur preferentially during phase 2 of SHM. Unexpectedly the mutation patterns resembled those seen with Poln and MSH2 deficiencies suggesting an additional pathway for A:T mutations conserved in mice and humans. SHM and CSR result in the formation of DNA strand breaks that in other cells could lead to cell cycle arrest driven by p21 and p53-d3pendent apoptosis. IRF8 also appears to collaborate with BCL6 in suppressing the activities of p53 and, secondarily, p21. Recently, we showed that IRF8 also activates the transcription of MDM2, a protein that normally induces degradation of p53. Suppression of p21, p53 and activation of MDM2 thus makes it possible for B cells to tolerate double stranded DNA breaks that occur physiologically during SHM and CSR. This feature of GC B cells might have he unfortunate consequence of allowing escape of self-reactive cells that could contribute to autoimmune diseases. Patients with multiple sclerosis regularly develop autoantibodies reactive with myelin basic protein (MBP) that may contribute to neuronal degradation. Remarkably, some of these autoantibodies were demonstrated to function as abzymes antibody enzymes with specificities that might make them useful for purposes of diagnosis and for following disease progression. Autoantibodies are also a prominent feature of the BXSB-Yaa mouse model of systemic lupus erythematosus. Previous studies showed that these mice express high levels of IL-21, a cytokine produced by follicular T helper cells known to drive GC reactions. To determine if IL-21 has a role in disease pathogenesis, the mice were made deficient in expression of the IL21 receptor (IL-21R). Remarkably, IL-21R-deficient BXSB-Yaa mice were essentially normal with none of the autoimmune or other features that normally characterize that strain. GCs are usually thought of a component of T cell-dependent (TD) responses to protein antigens. However, they can also occur as part of T-independent (TI) responses to polysaccharide antigens. Comparisons of gene expression patterns for GC arising from TD and TI responses revealed a TD gene signature that was unexpectedly comprised of multiple genes involved in the growth and guidance of axons. In keeping with this finding, human and mouse GC B cells were seen to form long neurite-like structures. This observation adds to the list of similarities between the neurological and immune systems and raise the possibility that antigen uptake, motility and process formation by B and dendritic cells may be driven in a neurite-like manner. Finally, we have been working with mice mutagenized with ENU to identify novel genes that induce leukocytosis or leukopenia. These mice were generated as part of a NHLBI program a the Jackson Laboratory to identify genes involved in cardiac, respiratory, or hematopoietic differentiation and function. We have identified the abnormalities associated with two mutations as affecting expression of IL7 and Tnf. These mutations effectively yielded null alleles in C57BL/6 mice. While conventional knockouts of these alleles were made previously, they do not exist on a pure inbred background or without the continued presence of neomycin selection genes that may confound phenotypes. These mice will permit clean assessments of the biology resulting from deficiency of either gene.
Showing the most recent 10 out of 57 publications
