During the development of multicellular organisms, precursor cells mature into specialized cell types, such as muscle cells or blood cells. This cellular differentiation must be carefully orchestrated to generate the correct numbers and types of cells at the right time and place. Unravelling the regulatory circuitry involved is a major goal of developmental biologists. In a striking illustration of Louis Pasteur's famous statement, “Chance favours only the prepared mind”, a fortuitous mutation in a mouse has led Kappes and colleagues to a breakthrough in understanding how immune-cell precursors adopt their appropriate cell fate (He et al.1, page 826 of this issue).

In the mammalian thymus, precursor cells called thymocytes give rise to two types of mature immune cell: CD4 helper T cells, which alert the immune system when the body is invaded by a pathogen and coordinate an inflammatory response; and CD8 killer T cells, which destroy cells that have been invaded. Each T cell has receptors for a specific antigen on its surface. The gene that encodes these receptors is rearranged during the cell's development to produce one of a huge potential variety of receptor genes. The range of receptors consequently produced by the entire population of mature T cells allows the immune system to recognize virtually any pathogen. These T cells probe other cells they come across, and respond when they find ‘foreign’ peptides that might signal the presence of a pathogen. The foreign peptides become bound to molecules of the major histocompatibility complex (MHC) and are displayed on the surface of invaded cells and other immune cells. The MHC comes in two varieties: class I, which is recognized by the CD8 killer T cells, and class II, which is recognized by the CD4 helper T cells. The CD8 and CD4 proteins aid the interaction between the MHC and the T-cell antigen receptor (Fig. 1).

Figure 1: The anatomy of a cell-fate decision.
figure 1

In the thymus, immune-cell precursors called thymocytes receive positive selection signals when their T-cell receptors (TCRs) recognize ‘self’ peptides (red circles) bound to molecules of the major histocompatibility complex (MHC) displayed on thymic epithelial cells. This interaction is helped by the CD4 or CD8 proteins. a, The thymocyte recognizes class II MHC, and receives a moderate, prolonged signal. This leads to production of Th-POK, a gene-regulatory protein that in turn controls other genes (including repression of the CD8 gene), causing the cell to adopt a CD4 fate. TCR signalling may also work independently of Th-POK (dashed line) to promote the differentiation or survival of CD4 cells. b, The thymocyte recognizes class I MHC, and receives a weak, transient TCR signal that fails to increase production of Th-POK. TCR-induced events in the absence of Th-POK lead to changes in gene expression (including repression of the CD4 gene), causing the cell to adopt the CD8 fate.

In the thymus, immature T cells express both CD4 and CD8. The cells test whether their newly formed antigen receptors can bind to MHC complexes on thymic epithelial cells that display ‘self’ peptides, derived from the body's own proteins. Most of the thymocytes fail the test and die. Remarkably, binding to the peptide–MHC complex not only leads to the survival of thymocytes, but also directs them to the appropriate lineage: recognition of class I MHC leads to development of a CD8 T cell, and recognition of class II MHC leads to development of a CD4 T cell (Fig. 1).

The work by Kappes and colleagues1 defines the molecular trail from the thymocyte–MHC interaction to the change in gene expression that guides the cell to its ultimate fate. While screening mice bred from two lines carrying different genetically engineered mutations, the authors noted2 that some of the progeny had an unusual characteristic that was not the result of either mutation — a complete absence of mature CD4 T cells. A spontaneous mutation, which they termed ‘helper deficient’, or HD, was responsible for the defect. Further analysis revealed that HD did not simply cause loss of CD4 cells, but rather produced a ‘cell-fate switch’ by which thymocytes that would normally have matured into CD4 T cells instead gave rise to CD8 T cells3.

Although there are a number of engineered mutations that can also redirect the CD4 versus CD8 fate decision (reviewed in refs 4, 5), most of these mutations lead to proteins that have abnormal functions (rather than causing disruption of a normal function), and thus they may not reflect the true functions of the unmutated genes. Furthermore, most of these engineered mutations cause only a fraction of thymocytes to make the wrong cell-fate choice. HD is the only mutation known that not only is a loss-of-function mutation, but also causes a wholesale conversion of thymocytes from the CD4 to the CD8 lineage, and so it seems to hold the key to understanding this cell-fate decision.

Seven years after the original description of the HD mutation, Kappes and colleagues1 have now identified the gene affected by it. It encodes a transcription factor called Th-POK (for T-helper-inducing POZ/ Krüppel factor). Transcription factors regulate gene activity, often in a tissue-specific manner. The discovery, therefore, that the HD mutation changes an amino acid of a transcription factor made perfect sense. The authors also show that, in the thymus, Th-POK is found specifically in mature CD4 T cells, and not in CD8 T cells. Moreover, enforced expression of the Th-POK gene in thymocytes transforms cells that would have been CD8 into CD4 cells, the opposite effect to the HD mutation. Thus, Th-POK is necessary and sufficient to induce the gene-expression programme for a CD4 cell fate and to suppress the CD8 programme.

Other transcription factors have been implicated in T-cell development, including GATA3, which is required for CD4 T-cell maturation6,7, and Runx3, which silences the gene encoding CD4 in CD8 T cells8. However, mutation of their corresponding genes does not cause a cell-fate switch, but merely impairs the development of one cell type. So Th-POK seems to be at the top of a gene-regulation hierarchy that controls T-cell fate.

Many tantalizing questions remain. What controls expression of the Th-POK gene in developing thymocytes? The observation that Th-POK synthesis is increased in thymocyte precursors that recognize class II MHC, but not class I MHC, implies that the gene's expression is regulated by engagement of class II MHC. This may be controlled by differences in T-cell receptor signalling on recognition of class I compared with class II MHC, as predicted by a popular model for CD4 versus CD8 lineage commitment (reviewed in refs 4, 5). Further analysis of the regulation of Th-POK by T-cell-receptor signalling will be needed to test this idea.

What genes are regulated directly by Th-POK? The gene-expression programmes of CD4 and CD8 cells differ in terms of the production of GATA3, perforin (a T-cell effector protein), and CD4 and CD8 themselves. However, it is not yet clear whether these differences in gene expression are direct or indirect consequences of Th-POK activity.

Finally, self-reinforcing feedback loops are often used to ‘lock in’ cell-fate decisions, and there are hints that Th-POK is part of such a regulatory circuit. HD mice expressing class II MHC, but not class I MHC, show increased synthesis of messenger RNA for Th-POK in thymocyte precursor cells. However, the redirected CD8 T cells that ultimately develop in these mice lack Th-POK. Whether the requirement for Th-POK function to maintain Th-POK gene expression involves auto-regulation of the gene or involves other components of the gene regulatory hierarchy remains to be seen.

The quest to understand the CD4 versus CD8 lineage decision has focused mostly on the question of whether MHC recognition instructs the cell-fate decision (the instructive model), or whether cells choose their fate randomly and are then tested for the presence of the appropriate CD4 or CD8 molecule (the stochastic or selection model)9,10. These models, although useful, assumed that T-cell fate determination occurs as a single discrete step, but it has become clear that it is a multi-step process involving feedback and reinforcement. The identification of Th-POK as a key in the T-cell fate decision should open the way to deeper mechanistic insights into how cells are guided to their destiny.