Our studies have addressed several major questions: 1) We have developed a novel method for the stabilization of Foxp3 expression in human Tregs upoin expansion in vitro. Adoptive cell therapy with expanded populations of Foxp3+ Treg has proven effective for the prevention of acute GVHD in mouse models and recent studies suggest that a similar approach will be effective in man. Most clinical applications of Treg therapy will require large numbers of cells. Expansion of even highly purified populations of hTreg frequently results in a loss of Foxp3 expression. Similar to our studies in the mouse, the expression of Helios has allowed us to define two populations of human Tregs. Foxp3+ Helios+ hTregs have a completely demethylated the TSDR (CNS2) region of the Foxp3 locus, whereas the TSDR of the Foxp3+Helios- cells is only 50% demethylated. One interpretation of this study is that the Foxp3+Helios- population is composed of 2 subpopulations, one of which expresses a fully methylated TSDR that can potentially lose Foxp3 expression on in vitro or in vivo expansion, whereas the second population may represent a population with greater stability. Addition of a TLR9 agonist ODN during the Treg expansion culture resulted in a markedly enhanced frequency of the Foxp3+Helios+ cells. The effects of the ODN were observed after 5 days of exposure but were also maintained for as long as an additional 16 days after the cells were washed and further expanded in the absence of the ODN. Further studies demonstrated that any random sequence ODN was capable of stabilizing Foxp3/Helios expression and that the effects of the ODN were not mediated by TLR9. As ODNs are already available for clinical use, it is likely that this protocol can readily be adapted for the preparation of clinical-grade Tregs. 2) We have analyzed in depth the Treg phenotype in a large series of patients with systemic lupus erythematosis (SLE). FoxP3 is not a reliable marker for distinguishing Treg cells in humans due to the fact that FoxP3 may be up-regulated in activated, conventional T (Tconv) cells. Recently, the marker CD45RA has been added to distinguish RA+, resting FoxP3+ Tregs from RA- activated FoxP3hi Tregs. This scheme has also included a third non-Treg group that is RA- and expresses lower levels of FoxP3 similar to that seen in RA+ Tregs. Previous data demonstrated that in normal individuals as well as in patients with SLE, the largest subset of FoxP3+ cells is, in fact, the non-Treg group. We have recently demonstrated that the transcription factor Helios is a marker of thymus-derived Tregs in both mouse and man. The major goal of this study was to determine if expression of Helios could reliably indicate what fraction of FoxP3+cells are true Tregs in healthy controls and in SLE patients. Samples included 35 healthy donors and 52 SLE patients (23 SLEDAI 0;19 SLEDAI 2-4;10 SLEDAI 6-20). Ficoll-purified PBMCs were surface stained followed by fixation/pemeabilization and intracellular staining prior to FACS analysis. When appropriate, cells were pre-stimulated for 4-5 hours with PMA/ionomycin/golgistop. CpG methylation analysis of sorted cells was performed using the Qiagen EpiTect platform. We found that CD4+ T cells in SLE patients contain both resting (RA+) and activated (RA-/FoxP3hi) Tregs and that the majority of cells in each group were Helios+, (3/4 and 4/5, respectively). Surprisingly, even within the sub-group defined as non-Tregs, the majority of the cells in both healthy controls (2/3) and SLE patients (3/4) were Helios+. All FoxP3+ Helios+ cells, including those within the non-Treg RA- subset, were found to be demethylated at the FoxP3 locus (a gold standard epigenetic mark of the Treg lineage), as compared to Helios- cells. In pre-stimulated cells, FoxP3+ Helios- cells consistently produced significantly high amounts of cytokines (IFN-gamma and IL-2), whereas FoxP3+Helios+ cells produced essentially no cytokines, which is characteristic of Tregs. In SLE patients with mild and highly active disease, there was a significant increase in both the % Foxp3+ Helios+ and % FoxP3+Helios- relative to both healthy controls and inactive patients. There was not a significant difference for absolute numbers of either Helios+ or Helios-cells, likely due to the significant reduction in total CD4 counts in more active patients. Thus, expression of Helios is a highly useful tool for distinguishing true Tregs from FoxP3+ cells that include activated Tconvs. The use of Helios has allowed us to de-convolute the largest subset of CD45RA-based grouping of FoxP3 cells. Furthermore, we have also shown that active SLE patients do have a higher frequency of FoxP3+Helios+ Tregs, but that in active SLE patients these are counter-balanced with a higher frequency of Foxp3+Helios- cells that contain cytokine-producing Tconvs. Future studies may make use of Helios to reliably monitor both true Tregs and activated Tconvs in SLE and other autoimmune diseases. 3. We have used RNA-Seq to analyze human Treg cell gene signatures. Microarray studies comparing hTregs or mTregs with Tconv (with or without prior activation) have provided insights to genes that are differentially expressed in Tregs, the so-called Treg signature. In an independent attempt (in collaboration with the Muljo lab) to find the unique molecules associated with hTregs, we are exploring sequencing of RNA-based libraries (RNA-seq) from human Treg in an attempt to provide a unique digital gene signature. This comprehensive approach guaranties an unbiased transcriptome analysis compared to existing microarray based queries for Tregs. The high sensitivity and accuracy of RNA-seq in detecting genes with very low or very high levels outperforms microarrays across many orders of expression magnitude. Using this approach we compared transcriptome signatures of human Tregs to conventional T cells at various phases of activation. We analyzed the RNA-seq reads using DEGseq, an R package that integrates three different methods (LRT, FET &MARS) to identify differentially expressed genes as well as by GO analysis to define cell membrane associated antigens. The differential expression of a panel of these gene has been validated further in multiple donors under different activation conditions at different time points using nanostring technology. Our results verified all known Treg signature genes including, but not limited to, FoxP3, IKZF2, IFNand CTLA4. A number of other thus far unknown Treg genes were identified and were validated as Treg-specific in multiple healthy donors and cord blood samples under different activation conditions. The most promising leads are being tested for their potential role in Treg suppressor function in gene knock down experiments in human Tregs.
Showing the most recent 10 out of 17 publications
