Conventionally, gc expression has been thought to be developmentally set and not affected by the differentiation or activation status of T cells. Our previous works have shown that this is not that case, and that gc expression is dynamically controlled during T cell development and homeostasis. Using newly established mouse models, such as the soluble gc (sgc) transgenic mice, we found that alteration in gc availability or function can control the generation of thymic invariant NKT cells (iNKT cells) and the differentiation of innate type CD8 T cells. Thus, it is not the expression of gc per se but also the quantity of gc surface expression that controls cytokine receptor signaling. In this regard, it was interesting to learn that immature CD4+CD8+ double-positive (DP) thymocytes express uniquely low amounts of gc compared to immature CD4-CD8- double-negative (DN) or CD4, CD8 single-positive (SP) thymocytes. To understand the molecular basis behind this observation, we addressed the role of the transcription factor RORgt which is uniquely and highly expressed in immature DP cells. We hypothesized that the specific expression of RORgt would be associated with suppression of gc expression in DP thymocytes. Thus, deletion of RORgt could possibly revert gc expression to the amounts found in DN or SP thymocytes. This was precisely the case, because we found that RORgt-deficient DP thymocyte upregulated gc surface expression to comparable amounts found on mature T cells. RORgt is known to induce expression of the anti-apoptotic factor Bcl-xL in immature DP cells. Consequently, we wished to know if Bcl-xL expression would be sufficient to suppress gc expression in immature DP cells. To this end, we generated RORgt-deficient mice that express a Bcl-xL transgene, and we found that Bcl-xL was sufficient to downregulate gc expression on DP thymocytes. These results reveal a new regulatory pathway of cytokine receptor expression that can be controlled by pro-survival factors, such as Bcl-xL. These results also suggest that surface cytokine receptor expression can be controlled in many layers and not simply by transcriptional control as conventionally accepted. In this regard, we have been also analyzing the role of cytokine receptor pre-association in gc cytokine signaling. Our previous work demonstrated a direct interaction of gc with other proprietary receptors of the gc family, such as IL-7Ra and IL-2Rb. Detailed analysis of the binding affinities between these cytokine receptor pairs revealed a dramatic preference for IL-7Ra/gc association compared to IL-2Rb/gc binding. If this would be the case, we predicted that all gc molecules would be sequestered and pre-bound to IL-7Ra, thus limiting its availability to other gc family receptors including IL-2Rb. Such a pre-association model of cytokine receptors puts forward a new layer of controlling cytokine signaling, and we found it important to demonstrate its role in T cell activation and differentiation. To this end, we used fluorescent beads to quantify the absolute number of gc cytokine receptors on cell surface of T cells. We found that IL-7Ra proteins outnumbered gc molecules at a ratio of four to one on surface on resting naive CD4 T cells. On the other hand, other gc cytokine receptors such as IL-4Ra or IL-2Rb were expressed at significantly lower numbers than gc. These results revealed that the availability of gc proteins is limited, and that IL-7 signaling is curtailed on CD4 T cells because of limiting amounts of gc and not IL-7Ra. To further understand the quantitative effects of cytokine receptor availability and signaling, we have now generated a series of new experimental models that include enforced expression of gc or IL-7Ra proteins on T cells. The downstream biological effects of altering gc availability is currently under investigation. Finally, to further understand the impact of alternative splicing on cytokine receptor expression, we focused on alternativesplicing of other receptors in the gc cytokine family. Because we found soluble IL-7Ra proteins in serum of both human and mice, we hypothesized a potential role for these protein products in controlling T cell immunity. In humans, soluble IL-7Ra has been previously described, and it is known that soluble IL-7Ra proteins are produced by alternative splicing of the IL-7Ra pre-mRNA. In mice, however, soluble IL-7Ra proteins have not been reported, and there is no molecular evidence available for an alternative IL-7Ra mRNA splice isoform in mice. Thus, our observation that mice do contain soluble IL-7Ra proteins in serum was surprising. In humans, soluble IL-7Ra is produced by alternative splicing that omits exon 6, which encodes the entire transmembrane region. However, we were unable to detect such alternative transcripts in mice, suggesting that the mechanism to produce soluble IL-7Ra proteins differ between humans and mice. In fact, cloning of IL-7Ra mRNA species from mouse T cells showed that alternative splicing of IL-7Ra pre-mRNA utilized a distinct mechanism from human T cells in that is use intron-retention, instead of exon exclusion, for alternative splicing. Whether this is the only mechanism to generate soluble IL-7Ra proteins need to be examined, and we do not exclude the possibility of membrane protein shedding to produce soluble IL-7Ra proteins. Whether such soluble IL-7Ra proteins are indeed involved in controlling T cell immunity and T cell differentiation is an important question that is currently under investigation.
Showing the most recent 10 out of 25 publications
