There are special moments in our lives as scientists when we come across a conceptual synthesis that sheds enormous light on our own research. I enjoyed such an illuminating experience when I discovered the elegant papers of Martien Kapsenberg and Pawel Kalinski on the role of dendritic cells (DCs) in immune polarization. Their work, in common with my own, addressed a core question in microbial immunity regarding how different infectious agents (such as bacteria, protozoa, worms and viruses) trigger distinct CD4+ T helper (TH) cell responses.

exposure to specific microbial products determines the differentiation of DCs to provide a specific type of signal 3

The prevailing view of TH cell differentiation at the start of the 21st century was that TH1 versus TH2 cell fate was determined by quantitatively and qualitatively distinct signals relating to T cell receptor (TCR) signal strength, co-stimulation or the cytokine milieu. However, the cellular source of these signals and how infectious agents might trigger them were poorly understood. Kapsenberg and Kalinski proposed a three-signal model of immune polarization in which these functions were all determined by one cell, the DC, through its distinct innate responses to different microorganisms. In their model, TH cell differentiation requires a third signal delivered by the DC, in addition to MHC–peptide–TCR ligation ('signal 1') and costimulatory molecule engagement ('signal 2'). Signal 3 was loosely defined and included other stimuli (such as prostaglandins), in addition to polarizing cytokines. Importantly, Kapsenberg and Kalinski introduced the concept of DC 'conditioning', whereby exposure to specific microbial products determines the differentiation of DCs to provide a specific type of signal 3. They provided evidence to support this principle using an in vitro model of human CD4+ T cell differentiation (J. Immunol., 2002).

The concept of DC conditioning as a crucial determinant of T cell fate has recently been called into question by data suggesting a more important role for pre-existing DC subsets that are 'hard-wired' to provide different types of signal 3. Although the evidence for this concept is compelling, we should not ignore the general plasticity of DCs and their ability to differentiate in distinct directions in response to specific pathogens. This is the concept so clearly formulated by Kapsenberg and Kalinski and supported by many of us who followed in their footsteps.