Children's Hospital of Pittsburgh Annual Progress Report: 2011 Formula Grant Reporting Period July 1, 2012 June 30, 2013 Formula Grant Overview The Children's Hospital of Pittsburgh received $228,401 in formula funds for the grant award period January 1, 2012 through December 31, 2013. Accomplishments for the reporting period are described below. Research Project 1: Project Title and Purpose Regulatory T Cells and Tolerance after Blood and Marrow Transplantation Tolerance after blood and marrow transplantation (BMT) is achieved eventually in most patients after 1-2 years post-bmt as they become independent of drugs to avoid rejection or graft-versus-host-disease (GVHD). Regulatory T Cells (Tregs) are known to be important in sustaining tolerance, however, there is a great gap of knowledge after BMT in humans regarding their activity in disease state (GVHD) compared to health (tolerance). In this project we will isolate and analyze Tregs from patients experiencing GVHD and contrast these to Tregs isolated from patients free of GVHD. Once functional prerequisites for tolerance are discovered, novel targeted therapies can be devised for those patients who suffer from GVHD. Anticipated Duration of Project 1/1 /2012 12/31/2013 Project Overview Tolerance after blood and marrow transplantation (BMT) is achieved eventually in most patients after 1-2 years post-bmt as they become independent of pharmacological agents to avoid rejection or graft-versus-host-disease (GVHD). The hallmark of tolerance is unresponsiveness between host and graft tissues in the absence of any immunosuppressive (IS) drugs. Regulatory T Cells (Treg) expressing CD25 and FOXP3 were identified over 10 years ago as critical players in sustaining tolerance. More Tregs in the transplant graft itself or in the blood of BMT recipients is associated with less GVHD. However, beyond these numerical associations there is a great gap of knowledge regarding their functional profile and features in GVHD compared to those patients without it. The proposed studies would discover new biological characteristics of Tregs that are essential for tolerance as they suppress the function of conventional T cells (Tcon). In this project we will develop new assays to analyze Treg cells from BMT patients and contrast the functional features of Tregs purified from patients with or without GVHD including those who never had GVHD. These studies may identify new biomarkers for the presence or absence of GVHD and should also identify specific features of Treg cells that are prerequisites Children s Hospital of Pittsburgh 2011 Formula Grant Page 1
for suppressing Tcon to induce tolerance. A new in vitro model could become a valuable tool to monitor other autoimmune diseases as well. We propose two specific aims. Aim 1. Enumerate Tregs from patients longitudinally and determine their T Cell Receptor (TCR) diversity by spectratyping after purification based on the expression of FOXP3 and Helios transcription factors. Contrast the profile of Tregs between BMT recipients with and without GVHD. Aim 2. Design new functional assays to characterize and quantitate the biological profile and suppressive capacity of purified Tregs in vitro as they are mixed with conventional T cells from the same BMT recipient. Following non-specific and transplant recipient specific activation of Tregs and Tcon in the same co-cultures to model the in vivo scenario we will analyze Tregs by qpcr for cytokine and homing receptors, costimulatory and other critical suppressor molecules while Tcon cells will be tested in parallel for evidence of receiving suppressive signals. Principal Investigator Paul Szabolcs, MD Chief, Division of Blood and Marrow Transplantation and Cellular Therapy Children s Hospital of Pittsburgh of UPMC University of Pittsburgh School of Medicine Rangos Research Center, Room 5125 4401 Penn Avenue Pittsburgh, PA 15224 Other Participating Researchers Xiaohua Chen, PhD Research Assistant Professor, Division of Blood and Marrow Transplantation and Cellular Therapy, CHP of UPMC, University of Pittsburgh Expected Research Outcomes and Benefits Graft-versus-host-disease (GVHD) is the most clinically significant immune mediated disease after allogeneic blood and marrow transplantation (BMT). If prevented or successfully treated, GVHD will be followed by tolerance characterized by the functional and peaceful co-existence of host and graft tissues in the absence of any immunosuppressive (IS) drugs. Despite prophylactic drug combinations, GVHD still develops in 20-60% of BMT recipients and can cause mortality in excess of 20% of all allogeneic transplant recipients notwithstanding significant morbidity and severe reduction in quality of life. There is preliminary evidence that rapid reconstitution of regulatory T Cells (Tregs) confers protection from GVHD in humans, nevertheless, there is a great gap of knowledge regarding their functional and biological features after transplant, beyond apparent numerical associations. In fact, the role of Tregs is far less well defined in human immune disorders compared to laboratory animal models. The proposed studies would identify and discover biological characteristics of Tregs that are essential for human transplant tolerance and thus new biomarkers of GVHD may be discovered. There are several centers that have developed ex vivo expansion platforms towards human clinical therapy with remarkably little functional knowledge of the manufactured products. Our studies could Children s Hospital of Pittsburgh 2011 Formula Grant Page 2
also aid to better define investigational products that are intended for human cell therapy trials. Importantly, a new in vitro laboratory model of testing Treg characteristics and function could become a valuable tool for researchers in the fields of rheumatology and gastroenterology to monitor disease activity and develop prognostic algorithms for autoimmune connective tissue disorders such as systemic lupus, Crohn s disease, rheumatoid arthritis, systemic sclerosis, diabetes, and others. Summary of Research Completed We are extremely grateful to the generous support to pursue our research goals as described in our two Aims below. We describe below the various aspects of our studies with the exciting new development on Treg cloning detailed in Aim I b below. Aim I. Aim 1A. Enumerate Treg cell distribution In our first interim report in July 2012, we presented our newly designed 8-color FACS panel to detect Treg phenotypes subsets. Since then, we have further optimized our FACS gating strategies to enhance detection of Treg sub-populations with the goal to correlate those with their functions. Figure 1 illustrates our current gating scheme. A P1 region identifies all lymphocytes (Figure 1a). Amongst P1 region cells, CD4+ cells were identified (P2) as shown in Figure 1b. Within this P2 region, putative Tregs were identified by gating on a CD25+ and CD127 dim subset as identifiable by the P3 region, see Figure 1c. The distribution of FoxP3 and Helios transcription factor expressing Tregs was then analyzed by quadratic gating, see Figure 1d. The distributions of naïve, central memory, and effector memory phenotype in each of the Treg subsets, as identified by their Foxp3/Helios expression, can be further subdivided by additional quadratic gating, see example for FoxP3+/Helios+ double positive Tregs in Figure 1e. The percentile to their parents of each population was applied, and thereby the influence from the variation of parent numbers was excluded in this analysis. To examine the state of Treg activation, we designed another 8-color FACS panel to include anti-ki67 (proliferation marker)/activated-caspase3 (apoptosis marker) Abs (Figure 1f). According to recent publications from Shimon Sakaguchi, we also added a combination of CD45RA vs Foxp3 to distinguish activated Tregs (region II) from resting Treg (region I) and Foxp3+ non-tregs (region III), see Figure 1g. With this new panel in place, we have characterized the distribution of Treg phenotypes in healthy donors. A total of seven healthy donors have been tested so far. An average of 8.8% CD4+CD25+CD127low Treg in CD4 population was observed. Notably, there were four populations carrying different Foxp3/Helios phenotypes (see Figure 1d) amongst the CD4+CD25+CD127low Treg, including single Helios+/FoxP3- (30.6%), double positive Foxp3+/Helios+ (39.3%), and single Foxp3+ /Helios- (8%) and double negative of Foxp3/Helios (22%). Circa 15% of CD4+CD25+CD127low Tregs and 13% of Foxp3+/Helios+ Tregs carried central memory phenotypes (CD45RO+CD62L+). Children s Hospital of Pittsburgh 2011 Formula Grant Page 3
Very recently, we started to monitor these phenotypes from patient samples. Pretransplant serotherapy by Alemtuzumab or ATG is routine in the allogeneic transplant population in our clinical practice; therefore we have a fairly low frequency of grade II or higher acute GvHD ( ~20%). Acute GVHD refers to GVHD that is diagnosed in the first ~ 100 days after transplant. With the relatively few patients, we have been able to test only one BMT recipient so far. As a comparison and control population, we have also examined cord blood recipients patients, some with mixed chimerism within the T cell fraction. A third population of patients, those with severe autoimmunity, was also enrolled on our IRB-approved protocol and tested. So far two cord blood recipients and three patients with autoimmunity have been studied. In comparison with healthy donors, there was no significant alteration in average numbers of CD4+CD25+CD127low (Figure 2A), single Foxp3+ (data not shown), and single Helios+ (data not shown) Tregs in the patients tested. However, we found, unexpectedly, a significant decrease of Foxp3+/Helios+ (double positive) Tregs carrying central memory phenotype (CD45RO+/CD62L+) in patients with autoimmunity, while this population showed a trend of increase in cord blood patients with mixed chimerism (Figure 2B). These results indicate that Foxp3+Helios+ central memory Tregs, rather than single Foxp3+ or Helios+ Tregs, may play an important role in the formation and maintenance of the immune tolerance. When we restrict the FACS analysis on the Sakaguchi regions (as defined by CD45RA/Foxp3 amongst CD4+/CD25 bright cells) and examine proliferation and apoptosis (Ki67/Caspase 3 respectively), we found that majority of Foxp3+Helios+ central memory Tregs (Figure 3A, 3B) belong to Sakaguchi region II -aka active Tregs which are CD45RA-/Foxp3 high (Q4-1 in Figure 3D). These cells have undergone significant proliferation, identifiable by the strong expression of Ki67 (see events in the Q1-1 quadrant in Figure 3C). When we examined the Treg profile of the subject with GVHD, we found little if any proliferation (low expression of Ki67 in Q1-2 Figure 4A) amongst Foxp3+/Helios+ Tregs, in comparison with that in a cord blood patient with mixed chimerism and in good clinical condition free of any GVHD, (Q1-2 Figure 4B). However, the overall total Treg numbers as identified by the CD4+CD25+CD127low phenotype were in normal range (data not shown). It is plausible that functional inefficacy of Foxp3+Helios+ Tregs and not their numerical paucity may be one reason for development of GVHD. Importantly, these new findings were detected directly from unmanipulated patient samples to reflect the human disease state without experimental influence leading to accidental bias. If the findings were confirmed with more cases, and Treg expansion would start with a defined phenotype it may lead to favorable response when treating GVHD by focusing on production of highly functional and stable Tregs. The findings on functionally and phenotypically distinct Tregs may also help to devise more powerful biomarker assays to predict the presence or absence of overall tolerance versus exaggerated/pathological immune responses (i.e., GVHD) post-hsct. Aim 1B. Treg single cell cloning Because Foxp3 and Helios genes are transcription factors it is not feasible to purify viable primary human Treg sub-populations by gating on these intracellular markers. Therefore, we Children s Hospital of Pittsburgh 2011 Formula Grant Page 4
have developed a novel single cell cloning technique with the goal to characterize Treg subpopulations. After expansion, we may sacrifice a fraction of the clone for characterization while many more are available for functional studies. The conventional method of T-cell cloning requires feeder cells to support single cell growth, which may cause contamination of unwanted cells or more likely lead to additional contact and soluble signals and might cause functional bias of Tregs expanding from single cells in this milieu. After months of pilot experiments, we have successfully established a new protocol for single Treg cloning without feeder cells (Figure 5), with a remarkable ~ 20% cloning efficiency for single Treg clones. To verify single cell clonality, we established TCR qpcr and TCR immunoscope for fast screening single cell cultures, another technical advance in this field. Figure 6 shows TCR qpcr profiles tested from the PBMC of a healthy donor (Figure 6A) or from the single cell cloning (Figure 6B); and TCR immunoscope profiles representing the PBMC of a healthy donor (Figure 6C) or the clonal progeny of a single cell (Figure 6D). Aim I C. TCR repertoire analysis Our original hypothesis was that the TCR repertoire of circulating Tregs from patients with GVHD would differ from that in patients not having experienced GVHD. The main obstacle to test this hypothesis has been the low cell numbers of Tregs that we have isolated, reflecting both CD4+ lymphopenia post transplant and the overall small volume of blood samples obtained from young children (4-8 ml range). Nevertheless, we have titrated the assay to the lowest cell number sufficient for efficient RNA purification. By adding carrier during the purification procedure, we can purify RNA from 1000 cells (data not shown). However, we have not yet obtained reproducibly clear signals on TCR immunoscope with RNA purified from 1000 cell. Aim II. After initially setting up a standard CFSE proliferation assay, we have performed further optimization to obtain differentiation peaks to give much distinct proliferation features. Figure 7 shows differentiation peaks generated at day 5 of a Tcon culture. We are currently titrating the dose of antigen presenting cells (APC), both live cells and artificial bead based signals, to determine the best conditions for in vitro Tcon activation to sufficiently quantitate the suppressive features from Treg. We have had surprising proliferative responses from Tregs in the presence of unnecessarily potent stimulatory signals. We are actively performing experiments to refine and enhance this functional assay with the inclusion of measuring the suppression of Th1 cytokine secretion of Tcon cells, particular IFNγ. In summary, by using our newly established 8-color FACS panels and gating strategy, we found preliminary evidence for the phenotypic shift of Foxp3+Helios+ central memory Tregs in the tested patient with acute GVHD (allogeneic responses), while the fraction of Tregs amongst CD4+ T cells remained in the normal range. In contrast, a trend in the increase of this T reg subset was seen in the two cord blood patients with mixed chimerism and no evidence of GVHD (allo-transplant tolerance). In the two tested cases of autoimmune disease, and no allogeneic cells in vivo, we found a significant decrease in the Foxp3+Helios+ central memory Tregs population. Children s Hospital of Pittsburgh 2011 Formula Grant Page 5
These findings suggest that Foxp3+Helios+ central memory Treg may play an important role in generation and maintain of immune tolerance. If confirmed, in sufficiently large patient groups, these are possibly exciting new findings. Importantly, the dataset was obtained directly from unmanipulated patient samples to accurately reflect the actual state of human disease such as GVHD and/or autoimmunity. If these very preliminary findings were to be confirmed in large datasets, it may guide the treatment of GVHD by focusing our attention on enhancing the number of functionally relevant Treg subsets rather than focusing on the composite total Treg pool. FIGURES Figure 1. FACS gating scheme for Treg phenotypes. Lymphocytes were identified in dotplots by FSC vs SSC (a); CD4+ cells were then gated (b). CD4+ T cells with CD25positivity and CD127 low expression were gated next (c). The distribution of Foxp3+ and Helios+ cells in CD4+CD25 high CD127 low Tregs were analyzed by quadratic gating (d). The distribution of CD45RO+/CD62L+ (central memory) versus other phenotypes in each Foxp3/Helios population were analyzed by additional quadratic gating (e); Ki67/Caspase 3 expression (f) and CD45R/Foxp3 (g) were also tested. Children s Hospital of Pittsburgh 2011 Formula Grant Page 6
Figure 2. The comparison of the distribution for CD4+CD25+CD127low population in CD4+ cells (A) or for Foxp3+Helios+ central memory Treg in CD4+CD25+CD127low cells (B) in healthy donors, in patients with autoimmunity or in patients with mixed chimerism and no evidence of autoimmunity or GVHD post-hsct. A B Figure 3. Characterization of Foxp3+Helios+Treg (A) with high expression of CD45RO+CD62L+ (B), and Ki67 (C); fell into Sakaguchi region II: CD45RA-Foxp3 high active Treg (D). A B C D Figure 4. Comparison of Ki67 expression of Foxp3+Helios+ Treg in a patient with agvhd (A) or with another healthy cord blood recipient in mixed chimerism (B) A B Children s Hospital of Pittsburgh 2011 Formula Grant Page 7
Figure 5. Single Treg clone expansion as reproduced by phase contrast inverted microscope at 7 days of culture (A) or at 12 days of culture (B). A B Figure 6. Identification of single cell clonality by TCRβ qpcr in single cell cloning (B) in comparison with PBMNC in a healthy donor (A); or by using TCR immunoscope in single cell cloning (D), in comparison with PBMNC obtained from a healthy volunteer donor (C). A B C D Children s Hospital of Pittsburgh 2011 Formula Grant Page 8
Figure 7. Differentiation peaks in CFSE proliferation assay are presented in dotplot (A) or in histogram (B). Children s Hospital of Pittsburgh 2011 Formula Grant Page 9