!Introduction
An exciting newcomer to the T helper (TH) cell field in the last few years is the TH-17 subset, which has been shown to play a role in antimicrobial immunity and in autoimmune and inflammatory pathologies in mouse models and is thought to be similarly involved in human disease (reviewed in refs. 1,2). The factors determining the differentiation of this subset have been clearly delineated in the mouse. Transforming growth factor-
(TGF-) and interleukin (IL)-6 are dominant in directing the differentiation of TH-17 cells from naive CD4+ T cells (reviewed in ref. 2), and the transcription factor ROR
T is an essential component in this process3(Fig.1). In humans, however, the situation has not been as clear. A number of papers have been published in the past year4, 5, 6, 7, 8 refuting the idea that the need for TGF- in the development of human IL-17–producing CD4+ T cells (TH-17). The verdict from these papers was that TGF- may be required for the differentiation of mouse TH-17 cells, but it is not required for human TH-17 cells, which instead make do with IL-6 and IL-1. But were these conclusions with respect to human TH-17 differentiation really justified? Two papers published in this issue of Nature Immunology by the Littman9 and Soumelis10 groups and a third in Nature11 now clearly demonstrate that, as in mouse, TGF- and RORT9 are indeed required for the differentiation of human TH-17 cells.
Figure 1: Cytokines can induce the development of helper cell subsets of effector T cells, including TH1 cells, by the action of STAT1, STAT4 and T-bet; TH2 cells, by the action of STAT6 and GATA-3; and the more recently described TH-17 subset, requiring the transcription factor ROR
t.
![Figure 1 : Cytokines can induce the development of helper cell subsets of effector T cells, including TH1 cells, by the action of STAT1, STAT4 and T-bet; TH2 cells, by the action of STAT6 and GATA-3; and the more recently described TH-17 subset, requiring the transcription factor ROR|[gamma]|t.](/ni/journal/v9/n6/thumbs/ni0608-588-F1.jpg)
TGF-, in concert with IL-6 produced by macrophages or dendritic cells, is a dominant factor in inducing TH-17 differentiation, together with a number of inflammatory cytokines, including IL-1, IL-21 and IL-23. In high doses or in the absence of these inflammatory cytokines, however, TGF- will instead induce development of Treg cells, expressing and dependent on Foxp3. APC, antigen-presenting cell; IFN, interferon; TNF, tumor necrosis factor.
The findings in these new studies elaborate and dissect in detail the regulation of TH-17 cell differentiation and are in keeping with findings reported in October 2007 at the International Cytokine Conference in San Francisco. Data was shown then from one of the groups that supernatants from dendritic cells stimulated with zymosan or -glucan induce the development of TH-17 cells, with a partial requirement for TGF- and IL-6 and an absolute requirement for IL-1 (ref. 12). And two of us have shown that TGF- together with IL-6 induces the development of TH-17 cells from naive human cord blood CD4+ T cells (B.S. and M.V., unpublished data). So why do these more recent studies demonstrate a critical role of TGF- and RORT in human TH-17 cell development, in keeping with mouse data?
One important consideration is that humans, in contrast to laboratory mice, are constantly exposed to a multitude of pathogens. It is therefore imperative to ensure that truly naive T cell precursors are isolated before one can make statements about de novo differentiation. Manel et al.9 went to great lengths to ensure this in their study by using highly purified naive human CD4+ T cells isolated from cord blood. Volpe et al.10 used CD4+ T cells purified by magnetic cell sorting from peripheral blood mononuclear cells for most of their study, but state that similar data are also obtained with purified CD4+ T cells isolated from human cord blood. The presence of activated/memory T cells producing other cytokines that inhibit TH-17 differentiation is a substantial obstacle in obtaining convincing TH-17 differentiation in vitro, and it may have restrained the effects of TGF- in the previous studies.
Researchers working in both mouse and human invariably use magnetic cell sorting to purify CD4+ T cells, which does not allow stringent isolation of only naive T cells. This may account for the previous difficulties in differentiating TH-17 cells from naive precursors encountered by researchers working in human systems, and it may also account for the low percentages of TH-17 cells often reported in mouse systems. Furthermore, contamination with antigen-presenting cells (APCs) may account for the apparent production of IL-6 by TH-17 cells, which has been suggested in some systems, but not others, by ELISA or RT-PCR analysis. Therefore, it is necessary to confirm the production of such cytokines by intracellular flow cytometric analysis to confirm that they are indeed products of TH-17 cells.
Another measure adopted by the researchers in these current publications is to carefully select their culture media: Manel et al.9 used serum-free medium to show that TGF- has an essential role in the differentiation of naive human CD4+ T cells toward the TH-17 lineage, similarly to results reported in mice. They further report that TGF- induced RORT expression but, paradoxically, inhibited its transcriptional activity, thus preventing expression of IL-17. This inhibition was relieved by the addition of a combination of inflammatory cytokines. Volpe et al.10 were able to induce significant TH-17 differentiation upon culture in serum-containing medium upon addition of TGF- plus inflammatory cytokines. However, it is of note that Volpe et al.10 used a specialized medium for their study (Yssel's medium) and a carefully selected fetal calf serum that may not contain consequential amounts of endogenous TGF-10.
In previous studies, the inability to observe TGF-–induced differentiation of TH-17 cells could be explained by the use of high serum concentrations, which may provide substantial amounts (several nanograms per milliliter) of latent TGF- that could become activated during the in vitro culture, especially if there are even small numbers of antigen-presenting cells present. Furthermore, contamination with platelets, which are the highest store of latent TGF- in the body, is difficult to avoid during isolation of peripheral blood mononuclear cells from human blood. It is also conceivable that human cells are more sensitive to small concentrations of TGF- than mouse cells are.
A large hurdle in differentiating TH-17 cells is that TGF- in the absence of inflammatory cytokines will induce Foxp3+ regulatory CD4+ T cell (Treg cell) development (reviewed in ref. 1). Even in the mouse, many researchers still use an excess of TGF-, compounded by extra TGF- in serum, and then report inhibition of various TH-17 parameters, such as IL-22 production. Instead, what is probably occurring is a preferential diversion to Treg cell formation at the expense of TH-17 differentiation. This could also account for the low percentages of TH-17 cells reported in many studies in the mouse. It has been reported that as little as 25 pg/ml of TGF- will still give rise to TH-17 cells2, and it appears that more than 2 ng/ml will begin to favor production of Treg cells, irrespective of the presence of IL-6 (ref. 13). Thus, more careful quantitation is essential to unraveling the complexities of TH-17 cell differentiation.
In previous studies in human systems, bone fide neutralizing antibodies to TGF- were not used to rule out the presence of endogenous TGF-. The present studies in this issue of Nature Immunology9, 10 and in Nature11 show that TGF- is required for both human and mouse TH-17 differentiation from naive CD4+ T cells. However, it is clear that the degree of polarization achieved falls short of what can be seen in the mouse system. The maximum percentage of human TH-17 cells obtained in all three studies is around 10%, which contrasts with several in mouse reporting that TGF- and IL-6 can induce 40 -
80% TH-17 cells under the right culture conditions. Furthermore, it remains unclear what the relative roles of inflammatory mediators such as IL-1, IL-6 and IL-23 are, and, in this regard, the conclusions reached in the studies by Manel et al.9 and Volpe et al.10 to some extent contradict each other. For instance, Volpe et al.10 show a requirement for IL-6 in human TH-17 differentiation, whereas Manel et al.9 do not, although IL-1, IL-21, IL-6 and IL-23 were all suggested to be involved in the process9, 10, 11.
On a more general level, the new findings raise a serious issue. The findings published last year that TGF- is not required for human TH-17 development were extrapolated in the Research Highlights section of Nature as suggesting that "mice are of limited use as models for the development of such cells in the human immune system"14. Such statements are unfortunate and unhelpful to the biomedical research community, as studies in both mouse models and human systems are critical for dissecting how the immune response is regulated. To progress to clinical trials, it is essential to have preclinical data obtained from in vivo mouse models and from in vitro work in both mouse and human. Clearly, mouse models of disease may not always show features identical to those of human disease, but it is important to compare mouse and human parameters and, when they do indeed differ, try to use this to advance scientific research instead of turning it into a competitive issue.
Many previously suggested apparent differences in the regulation of the immune response in mouse and human by cytokines have arisen, owing to different sources of cellular preparations used in the two species and to the purity and/or state of activation or naivety of the immune cells. It is of note that the identification of TH1 and TH2 cell subsets in human initially met with difficulties, but it eventually became clear that human TH1 and TH2 cells follow similar rules as their mouse counterparts (reviewed in ref. 15).
The studies in this issue of Nature Immunology9, 10 and in Nature11 'reinstate' TGF- as a critical factor in inducing both human and mouse TH-17 differentiation from naive CD4+ T cells, in concert with inflammatory mediators including IL-1, IL-21, IL-6 and IL-23. Future studies will undoubtedly optimize and fine-tune our understanding of the developmental program that shapes this interesting new T cell subset in both mouse and man.
