Only genuine CD4⁺CD25⁺ T_regs may be friends or foes

While we agree with many of the points raised by Fietta and Meloni, we also fear that they misinterpret some of our conclusions. Prior murine studies suggested that coexpression of CD4 and CD25 alone could demarcate a regulatory T-cell (T_reg) subset capable of suppressing graft-versus-host disease (GVHD) after mismatched stem cell transplantation (SCT).^1,2  In contrast to findings of murine studies, we found that increased frequencies of CD4⁺CD25⁺ or CD8⁺CD25⁺ donor graft T cells were positively associated with recipient GVHD in humans.³  We quantitated CD25 expression on donor CD8⁺ T cells as a control population, and not because we expected regulatory activity to be associated with the CD8⁺CD25⁺ subset.³  Because we found that the infusion of increased numbers of CD25⁺ T cells was positively associated with the risk of GVHD, we concluded that CD25 expression might have reflected activation within donor T cells.³  These data confirmed that the CD4⁺CD25⁺ phenotype should not be equated with T_reg function. However, our results in no way diminish the important work of others who have demonstrated the very real potential of T_regs to modulate human immune responses in a number of disease settings.

Fietta and Meloni cite a study by Ng et al⁴  as evidence that “the lack of expression of a phenotype of recently activated T cells should be always assessed on presumed T_reg subset.” While Ng et al demonstrated that cells stimulated with phytohemagglutinin for short periods expressed higher levels of CD69 than “naturally occurring CD4⁺CD25⁺ cells” in healthy human donors, it is notable that they also demonstrated that cells isolated purely on the basis of a CD4⁺CD25⁺ phenotype acted as functional T_regs, remaining hyporesponsive to mitogens and proving capable of suppressing activation in CD4⁺CD25^- T cells.⁴  While CD69 expression is closely associated with cytokine production after short-term stimulation⁵  and might help us to identify the subset of CD4⁺CD25⁺ T cells that are activated, persistently activated T cells may down-regulate surface CD69 expression, suggesting that the lack of CD69 expression might not be sufficient to demarcate T_regs. While it has been suggested that the intensity of CD25 expression may discriminate activated versus regulatory CD4⁺CD25⁺ T cells, we could not confirm this in our study, as we found that increased frequencies of CD4⁺CD25^high cells within donor grafts were also associated with recipient GVHD.⁶ 

Fietta and Meloni suggest that they would have conducted our study differently, by examining additional phenotypic or functional markers of T_regs, or by assessing Foxp3 expression within these grafts. Many of the studies they suggest (eg, functional studies demonstrating the ability of CD4⁺CD25⁺ T cells to suppress activation of CD4⁺CD25^- cells) could not have been performed given the limited number of total cells available within the convenience samples we used for our analysis. We began our investigation prior to recent studies establishing Foxp3 as a molecular marker of T_regs.⁷  Even now, there are no standard methods for quantitating T_regs on the basis of Foxp3 expression within a heterogeneous cellular population. In an essay that is nearly 35 years old, Ziman argued that “a typical scientific paper has never pretended to be more than another little piece in a larger jigsaw—not significant in itself but as an element in a grander scheme.”⁸  In this spirit, we hope that our study inspires other investigations that will leave off where we began, so that we may better understand the cellular allies and adversaries influencing both positive and negative posttransplantation outcomes.