Antigen-primed B cells form germinal centres, where, helped by follicular helper T (TFH) cells, they develop into high-affinity plasma and memory B cells. This process is under stringent control to avoid uncontrolled B cell proliferation and production of autoantibodies. Now, two studies published in Nature Medicine report that germinal centre reactions are controlled by regulatory T (TReg) cells with a TFH cell-like phenotype.

The two studies identified forkhead box P3 (FOXP3)+ TReg cells expressing the TFH cell-associated factors CXC-chemokine receptor 5 (CXCR5) and B cell lymphoma 6 (BCL-6) in the germinal centres of immunized mice. These follicular TReg cells expressed both TFH cell-associated and TReg cell-associated genes and were suppressive in vitro.

So, do follicular TReg cells develop from TFH or TReg cells? Adoptive transfer experiments indicated that follicular TReg cells originate from TReg cells. Moreover, Linterman et al. showed that thymus-derived but not induced TReg cells are the follicular TReg cell progenitors. In addition, they observed that, similarly to TFH cells, follicular TReg cells require signalling through CD28 and SLAM-associated protein (SAP) for their generation, and both studies demonstrated that BCL-6 expression is essential for their differentiation. Thus, BCL-6 expression and SAP-mediated signalling lead to the development of follicular TReg cells from natural TReg cells.

But what is the function of follicular TReg cells? To address this, Linterman et al. generated chimeric mice that contained SAP-deficient T cells (which cannot differentiate into TFH or follicular TReg cells), as well as T cells expressing the diphtheria toxin receptor under the control of the Foxp3 promoter. Selective depletion of follicular TReg cells with diphtheria toxin resulted in increased numbers of TFH cells and germinal centre B cells in response to immunization. Intriguingly, the numbers of non-antigen-specific B cells were increased, whereas the percentages of antigen-specific plasma cells and memory B cells were reduced in chimeric mice lacking follicular TReg cells, compared with the numbers in control chimaeras. Thus, the authors suggest that follicular TReg cells limit TFH cell numbers, and possibly select against those that are self-reactive, thereby providing a competitive advantage to high-affinity antigen-specific B cells.

By contrast, Chung et al. observed no increase in TFH cell numbers in response to immunization when they transferred naive CD4+ T cells together with BCL-6- or CXCR5-deficient TReg cells (which cannot differentiate into follicular TReg cells) into T cell-deficient recipients. In their model, the absence of follicular TReg cells resulted in increased numbers of germinal centre B cells but also in elevated levels of antigen-specific B cells and high-affinity immunoglobulins.

Taken together, the two studies describe the BCL-6-dependent differentiation of natural TReg cells into follicular TReg cells, which negatively regulate germinal centre B cell numbers. The seemingly opposing conclusions on the effect of follicular TReg cells on antigen-specific B cell responses may rely on the experimental approaches used, and further research will clarify the contribution of follicular TReg cells to the regulation of germinal centre reactions.