Whether the characteristics of a dendritic cell (DC)–CD4+ T-cell interaction determine whether the CD4+ T cell is primed or tolerized remains an open question. However, new insight into this is provided by two groups who have shown in situ that there are only subtle differences in the behaviour of CD4+ T cells after antigen encounter under conditions that result in priming or tolerization.

Previous two-photon microscopy studies that visualized T cells interacting with antigen-loaded DCs in situ indicated that, under priming conditions, antigen-specific T cells form stable contacts with DCs, whereas under tolerizing conditions, these stable contacts do not form. This has led to the hypothesis that the stability of the DC–T-cell interaction determines whether T cells are tolerized or primed. Because this hypothesis remains controversial, both groups set out to visually compare CD4+ T-cell priming and tolerance induction in situ.

To deliver antigen to DCs in vivo, Shakhar et al. used a fusion protein consisting of a DC-specific antibody fused to the ovalbumin (OVA) peptide that is recognized by T cells expressing the OT-II T-cell receptor (TCR). Antigen was delivered alone or in combination with a CD40-specific antibody to create tolerizing or priming conditions, respectively. T cells expressing both the OT-II TCR and enhanced green fluorescent protein (EGFP) were transferred to the immunized animals, and the behaviour of the cells in the lymph nodes was tracked intravitally. Enhanced cyan fluorescent protein (ECFP)-expressing, antigen non-specific CD4+ T cells were co-transferred with the EGFP+OT-II TCR+ T cells to distinguish antigen-specific and non-specific behaviour. During the first 6 hours of imaging, under both tolerizing and priming conditions, EGFP+OT-II TCR+ T cells moved more slowly than the antigen non-specific ECFP+CD4+ T cells. In addition, a large proportion of the EGFP+OT-II TCR+ T cells were immobile, spending more time arrested than the control ECFP+CD4+ T cells. The only differences between tolerizing and priming conditions were observed between 6 and 12 hours, when the EGFP+OT-II TCR+ T cells regained speed more quickly under tolerizing conditions. Subsequently, between 12 and 18 hours, there was again little difference in the speed of the cells and in the amount of time spent arrested between EGFP+OT-II TCR+ T cells and control ECFP+CD4+ T cells under either tolerizing or priming conditions.

By contrast, Zinselmeyer et al. induced priming or tolerance by oral administration of OVA in the presence or absence of cholera-toxin adjuvant, and they imaged mucosal and systemic lymph nodes both ex vivo and intravitally. In these studies, mice had previously been administered CFSE (5,6-carboxyfluorescein diacetate succinimidyl ester)-labelled CD4+ T cells expressing the OVA-peptide-specific DO11.10 TCR. Similar to the observations of Shakhar et al., under both tolerizing and priming conditions, DO11.10 TCR+ cells slowed down, formed clusters and appeared to stop. Further analysis indicated that the only differences were subtle: there were fewer T cells in each cluster under tolerizing conditions than under priming conditions, although the proportion of T cells entering clusters was greater under tolerizing conditions (particularly 20 hours after antigen feeding); also, at 20 hours after antigen feeding, the size of the clusters was greater under priming conditions.

These reports indicate that there are no marked differences in the initial phases of DC–CD4+ T-cell interactions that lead to the distinct outcomes of tolerance and priming. Indeed, stable interactions of antigen-specific T cells with DCs (which are thought to be the in vivo counterparts of immunological synapses) were correlated with activation and proliferation but not with tolerance or priming. However, further work will be required to determine whether the subtle differences that were observed in these studies are responsible for determining tolerance versus priming, as suggested by Zinselmeyer et al., or whether events at time points later than those studied in these reports control this commitment step, as suggested by Shakhar et al.