In this issue of Blood, Guo and colleagues investigate the origins of 2 components of the immune system, T cells and NK cells, and find that the formation of both lineages depends exquisitely on the activity of core binding factors.
Developmental immunologists have long sought to define the origin and genetic program of the cell that first seeds the thymus. Our current understanding is that the arrival of such an early T-lineage progenitor (ETP) in the thymus results in the rapid loss of B-cell potential (reviewed by Bhandoola et al1 ). This process is dependent on Notch signaling, as in its absence, thymic progenitors differentiate into B cells.2 The ETP, while having a bias to form T-cell progeny, retains the capacity to differentiate into natural killer (NK), dendritic, and myeloid cells.1 Survival and expansion of early thymocytes also depends on the cytokine IL-7. Despite the often low-level expression of many T cell–associated genes, the ETPs and the subsequent stage, termed double-negative 2 (DN2), are not irreversibly committed to the T-cell lineage; this is a process that occurs in DN3 cells.3
A number of transcription factors have also been implicated in early thymopoiesis, including the core binding factors (CBFs).3 CBFs consist of a complex of the DNA-binding Runx proteins (Runx1-3) and the cofactor CBFβ, and are required for many aspects of hematopoiesis, including stem-cell specification and megakaryocyte and lymphocyte development.4 CBFs are also the target of both sporadic and familial mutation in acute leukemia.5 In the current study, Guo and colleagues have made use of mice harboring a hypomorphic Cbfb allele, which, in conjunction with a null allele, reduces CBFβ levels to 15% of normal,6 thereby showing that CBFβ is essential for T-cell lineage specification. Interestingly, while some cells exist that have the appearance of thymic progenitors, they are markedly impaired in their ability to proliferate in a coculture system with stromal cells and the Notch ligand, delta-like 1 (OP9-DL1)—conditions that promote T-cell specification and later differentiation. The CBFβ mutant cells express virtually no T cell–specific transcripts and lack T-cell receptors. These elegant experiments clearly demonstrate that CBFs are critical for T-cell specification.
An important component to understanding early T-cell development is to determine how externally derived signals such as Notch and IL-7 are integrated with the intrinsic transcriptional networks in the progenitors themselves to promote cell-fate choices. Guo and colleagues have made use of the powerful OP9-DL1 system to address this issue, and demonstrate that although the mutant progenitors can respond to cytokines and Notch ligands, they do not appropriately integrate these signals to promote T-cell fate specification. These experiments suggest that CBFs are required to support a progenitor that then becomes Notch responsive, a finding that contrasts with hematopoietic stem cells in which the expression of the CBF component Runx1 is Notch dependent.7,8
Although the residual level of CBFβ appears to support relatively normal B-cell development, Guo and colleagues also found that mutant progenitors are incapable of differentiating into mature NK cells. NK-cell development depends absolutely on signaling through the IL-15 pathway, and in keeping with this, CBFs are required for CD122 (Il2rb, a component of IL-15R) expression.
Altogether, these studies place CBFs at the center of early lymphopoiesis; whether CBFs are first required in ETPs or prior to thymic colonization remains to be determined. There is evidence for common NK/T progenitors in fetal and adult hematopoiesis,1 but distinct deficiencies in both lineages remains a possibility to be addressed in future studies. Similarly, whether similar mutations in CBF components are responsible for the subset of severe combined immune deficiency (SCID) patients with normal B-cell counts who do not harbor mutations in the known components of the common γ chain cytokine signaling pathway remains to be investigated, although familial mutations in RUNX1 have not been reported to result in SCID.5
Conflict-of-interest disclosure: The author declares no competing financial interests. ■