Megan Levings

Regulatory T cell therapy has shown promise in treating autoimmune disorders, transplant rejection and graft-versus-host disease in early clinical trials. However, efficient manufacturing of clinical grade cells is still a significant hurdle that must be overcome before these therapies can see widespread use. Previous work showed that large numbers of pure, naïve Tregs can be isolated from pediatric thymus. This research aims to investigate the variables governing Treg expansion with serum-free media and non-cell-based activation reagents to develop manufacturing protocols to produce therapeutic doses of thymus-derived Tregs. First, we tested activation reagents, cell culture media, restimulation timing, and cryopreservation to develop good manufacturing practice compatible protocols to expand and cryopreserve Tregs. Cryopreservation tests revealed a critical effect of timing: only cells cryopreserved 1-3 days, but not > 3 days, after restimulation maintained high viability and FOXP3 expression upon thawing. We next investigated how changing cell density and feed frequency influenced Treg expansion, viability, and phenotype in a 3-week expansion protocol and found that Treg viability and expansion were correlated with the cell density at restimulation. Tregs restimulated at low cell densities (1x10⁵ cells/cm²) initially had high growth rates, viability, and FOXP3 expression, but at later culture times these parameters were reduced compared to slower growing Tregs restimulated at higher cell densities (5x10⁵ cells/cm²). High density expansion was associated with lower nutrient concentrations and higher accumulations of lactate, but this could be alleviated by decreasing the interval between feeds. We tested platforms to scale up Treg manufacturing and observed that Tregs expanded in gas permeable cell expansion bags were of higher quality than those expanded in agitated suspension culture or the G-Rex. Finally, we tested labelling expanded Tregs with a ¹⁹F-perfluorocarbon (PFC) nanoemulsion to enable in vivo tracking using MRI. While Tregs could be labelled with the ¹⁹F-PFC and detected in vivo in immunocompromised mice, labelling during expansion reduced cell viability, particularly after cryopreservation. Together, this research developed protocols and process understanding necessary to efficiently produce clinical grade Tregs, laying the groundwork for the first clinical trial of thymic Treg cell therapy in Canada.

View record

CD4⁺FOXP3⁺ regulatory T cells (Tregs) control inflammation and enforce self-tolerance to maintain immune homeostasis. Adoptive Treg cell therapy holds promise for a variety of immune-mediated conditions, including transplant rejection and autoimmune disease. The last decade has seen an increased appreciation for Treg heterogeneity and hence an untapped potential for therapeutic Tregs with tailored functional profiles. We aimed to enable the development of next-generation Treg cell therapies by investigating novel Treg functions, optimizing gene editing tools, and interrogating the parameters essential for optimal Treg stability and efficacy. First, we examined the potential of human Tregs to directly mediate tissue repair. In contrast to mouse models, human Tregs did not express the IL-33 receptor ST2, but Tregs engineered to overexpress ST2 exhibited several tissue-reparative features, including amphiregulin expression and a heightened ability to induce M2-like macrophages. Second, we optimized a CRISPR-based approach to knock-out/knock-in genes in human Tregs. Using this method, we found that the master transcription factor FOXP3 had subset-specific roles in regulating mature Treg function and a limited role in maintaining lineage identity. Finally, we applied genome editing in human Tregs to interrogate the role of the transcription factor Helios, implicated in mouse Treg stability. Human Tregs that spontaneously downregulated Helios expression during in vitro expansion exhibited phenotypic and functional instability, but CRISPR-based Helios-knockout Tregs did not exhibit discernable functional defects. Overall, this work highlights important species-specific considerations in the translation of biology to therapy and offers a method to explore future mechanistic questions in human Tregs. 

View record

Achieving transplant tolerance with regulatory T cell (Treg) adoptive immunotherapy is currently under investigation as a therapy to reduce graft rejection, improve long-term outcomes, and patient quality of life. Initial approaches involve expansion of naturally-occurring Tregs, either polyclonal or antigen-specific, however both of these approaches have several technical limitations that restrict implementation at a large-scale. To circumvent these limitations, this work describes an alternate approach to generate antigen-specific Tregs by expressing a chimeric antigen receptor specific for HLA-A*02:01 (A2-CAR), which activates Tregs in the presence of HLA-A*02:01, a tissue antigen allele that is commonly mismatched between transplant donor and recipient. In the first CAR Treg studies, the antigen-binding region (scFv) of the A2-CAR was derived from the mouse BB7.2 hybridoma, which could cause immunogenic responses and limit its efficacy in humans. Additionally, most CAR Treg studies to date employ CD28 and CD3 signaling domains to activate the cell, but alternative co-receptor signaling moieties have not been adequately tested. Two major improvements to CAR Treg technology are explored: (1) the scFv is humanized to reduce the immunogenicity of the CAR construct itself, rendering it less likely to cause immune responses in humans and (2) a collection of a variety of co-receptor intracellular domains are tested in place of CD28 to determine whether alternative signals can bestow Tregs with more beneficial functional properties. In the final chapter, a method for staining FOXP3, the Treg master transcription factor, using mass cytometry is described to enable thorough tracking of FOXP3⁺ Tregs and the rest of the immune compartment in patient samples from various tissues. Collectively, this body of work furthers our understanding of Treg immunotherapies and provides further support for their use in transplant settings.

View record

Adipose tissue (AT)-resident regulatory T cells (Tregs) are important regulators of local and systemic inflammation and metabolism. We have previously found that insulin receptor signaling inhibits Treg suppressive function in vitro, and hyperinsulinemia is associated with alterations in visceral AT Tregs in vivo. To directly test the role of Treg-intrinsic insulin receptor signaling, we generated Foxp3(cre)InsR(fl/fl) mice and fed them either chow or high-fat diet (HFD) to induce hyperinsulinemia and obesity. Compared to Foxp3(cre) mice, Foxp3(cre)InsR(fl/fl) mice had improved glucose tolerance and insulin sensitivity after 13 weeks of HFD. This protective metabolic effect was associated with increased brown AT ST2+ Treg numbers and functions and dampened AT inflammation in Foxp3(cre)InsR(fl/fl) mice, rather than changes in food consumption or energy expenditure. Moreover, Foxp3(cre)InsR(fl/fl) mice were protected from age-associated glucose intolerance and insulin resistance as determined in 52-week-old mice. Unlike in the HFD cohort, visceral AT Treg numbers and functions were greatly reduced in aged Foxp3(cre)InsR(fl/fl) mice, leading to increased inflammation in the visceral AT. Surprisingly, elevated AT inflammation was associated with improved metabolic outcomes. Gene expression analysis revealed differential gene signatures in AT Tregs and total AT inflammation, suggesting that the underlying mechanisms contributing to the metabolic syndromes were distinct between the two insulin resistance models. Together, these data suggest that hyperinsulinemia contributes to metabolic syndrome in part by differentially affecting Treg control of AT inflammation in diet-induced versus age-associated metabolic syndrome. In mice, lean visceral AT is populated with anti-inflammatory cells, notably Tregs expressing the IL-33 receptor ST2. Conversely, obese AT contains fewer Tregs and more pro-inflammatory cells. In humans, however, there is limited evidence for a similar pattern of obesity-associated immunomodulation. We used flow cytometry and mRNA quantification to characterize human omental AT in 29 obese individuals, 18 of whom had T2D. Omental AT Treg frequencies negatively correlated to BMI but were comparable between T2D and non-T2D individuals. Compared to human thymic Tregs, omental AT Tregs had a distinct gene signature. ST2, however was not detectable on omental AT Tregs from lean or obese subjects.

View record

Cell-based therapy with CD4⁺FOXP3⁺ Regulatory T cells (Tregs) has the potential to be an effective treatment to limit organ transplant rejection, graft-versus-host disease and autoimmunity. Yet, challenges for the successful implementation of Treg therapy include difficulties in isolating and expanding homogeneous cell populations while preserving their suppressive function. Further, to be effective, Tregs may need to express homing receptors for migration to inflammatory or tissue sites. The aims of this research were to investigate an alternate source of therapeutic Tregs, to test whether Treg homing capacity could be tailored, and to investigate the migratory capacity of engineered antigen-specific Tregs. To evaluate pediatric thymus – routinely removed during cardiac surgery – as a source of therapeutic Tregs, I isolated thymic Tregs as CD4⁺CD25⁺ cells using different protocols. Large numbers of thymic Tregs could be isolated and expanded in vitro to clinically relevant numbers while maintaining high FOXP3-expression. Thymic Tregs did not secrete pro-inflammatory cytokines and were suppressive in vitro and in a humanized mouse model. Next, I tested whether the addition of cytokines or metabolites to the expansion culture could tailor the homing capacity of thymic Tregs towards inflammatory and specific tissue sites. Addition of IFN-gamma and IL-12 yielded CXCR3⁺TBET⁺FOXP3⁺ Th1-Tregs with migratory capacity to Th1-chemokines in vitro while maintaining their Treg phenotype. Adding retinoic acid to cultures induced expression of the gut-homing receptors alpha4beta7-integrin and CCR9 on Tregs and elevated their gut-relevant suppressive mechanisms. Homing-receptor-tailored Tregs were epigenetically stable even after long-term exposure to inflammatory conditions and they were suppressive in vivo. Finally, I investigated how engineered antigen-specificity affected trafficking of Tregs in vivo, by transducing them with a chimeric antigen receptor (CAR) targeting HLA-A2. I tracked homing of luciferase-transduced HLA-A2-CAR Tregs longitudinally in vivo and found that HLA-A2-CAR Tregs migrated to and proliferated at HLA-A2-expressing skin grafts and could migrate to draining lymph nodes. Tregs recovered from lymph nodes maintained FOXP3- and CAR-expression. Collectively, these studies show that thymic Tregs could be used as an ‘off-the-shelf’ cellular therapy and tailoring Treg homing capacity offers a new and clinically-applicable approach to improving the potency and specificity of Treg therapy.

View record

Obesity is associated with chronic low-grade inflammation in visceral adipose tissue (VAT), which promotes the development of insulin resistance. The role of adaptive immunity in VAT inflammation has only recently been investigated. Initial studies suggest that VAT-resident regulatory T cells (Tregs) have a prominent role in suppressing VAT inflammation and correcting metabolic dysfunction in obese mice. I sought to investigate how Tregs in the VAT are regulated.Obesity is accompanied by a rise in insulin levels, and whether this hyperinsulinemia affects the progression of inflammation is not known. I first found that Tregs express the insulin receptor, and high levels of insulin inhibited IL-10 production and the ability of Tregs to suppress macrophages. In parallel, Tregs from the VAT of obese mice showed a similar decrease in IL-10 production, suggesting that hyperinsulinemia may contribute to the development of obesity-associated inflammation via an effect of insulin on Treg function. I then found that the majority of IL-10-expressing Tregs in the VAT expressed the ST2 chain of the IL-33 receptor. The proportion of ST2+ Tregs in VAT was severely diminished in obese mice, and this effect could be completely reversed by treatment with IL-33. IL-33 treatment also reversed VAT inflammation in obese mice, and resulted in a reduction of hyperinsulinemia and insulin resistance. These data suggested that IL-33 is critical for the maintenance of ST2+ Tregs in the VAT, and that delivery of IL-33 may be a new therapeutic approach to reverse obesity-associated Treg deficiency, inflammation and insulin resistance. It is not known whether Tregs and adipocytes can directly interact in the VAT. I found that soluble factors produced by adipocytes significantly increased survival and IL-10 production from Tregs, and caused a shift towards oxidative metabolism in vitro. Similarly, Tregs resident in mouse VAT have substantially increased expression of IL-10 compared to those found in the periphery, indicating that the interaction between Tregs and adipocytes may contribute to the functional specialization of Tregs in the VAT. Taken together, these data suggest that insulin, IL-33 and adipocyte-produced factors regulate IL-10-expressing Tregs and their ability to control inflammation in the VAT.

View record

During chronic inflammation and tissue injuries, various danger-associated molecules can be released and are able to potentiate inflammation and T cell responses. Among the many possible danger signals, I focused on studying high concentrations of extracellular ATP since it has been implicated in a variety of autoinflammatory diseases. ATP activates the inflammasome in macrophages, stimulates dendritic cell (DC) maturation, and inhibits regulatory T cell (Treg) function. However, how ATP regulates Toll-like receptor (TLR) responses in intestinal epithelial cells (IECs), which represent the front line of enteric defense, remains unclear. Therefore, I examined how ATP modulates TLR responses in IECs and found that it enhanced the response of IECs to a TLR1/2 ligand Pam₃CSK₄ primarily through the P2X7 purinergic receptor, leading to increased DC maturation and antigen-specific T cell proliferation. Furthermore, intra-rectal delivery of ATP lowered the activation threshold of epithelial cells to endogenous TLR ligands, making IECs more prone to immune activation. Since ATP is an important molecule that can potentiate inflammatory responses, the second aim of the study was to investigate if Tregs, including Foxp3+ Tregs and IL-10 producing Tr1 cells, can regulate ATP induced inflammasome activation and IL-1β production. I found that Tr1 cells inhibited the production of Il1b mRNA, inflammasome-mediated activation of caspase-1, and secretion of mature IL-1β, in an IL-10 dependent manner. Surprisingly, Foxp3+ Tregs, despite the production of IL-10, failed to inhibit IL-1β production. The important role of IL-10 in regulating inflammasome activation was further illustrated in the monosodium urate induced peritonitis model, where IL-10R-deficient mice had an increased influx of peritoneal neutrophils compared to wild type mice. Moreover, IL-1β production from macrophages derived from Nlrp3A350V knock-in mice, which carry a mutation found in cryopyrin associated periodic syndrome patients, was suppressed by Tr1 cells, but not Foxp3+ Tregs. Using an adoptive transfer model, I found Tr1 cells can protect against weight loss in mice expressing a gain-of-function mutation in NLRP3. Collectively, these data demonstrated the complex regulation of host response to cellular stress signal ATP, and that IL-10 producing Tr1 cells may have unique therapeutic effects in controlling ATP-mediated inflammasome activation via IL-10-mediated suppression.

View record

CD4⁺FOXP3⁺ T regulatory cells (Tregs) are potent suppressors of inflammatory immune activity. Cellular therapy with Tregs is a promising way to induce antigen specific tolerance in transplantation and autoimmunity, as it would allow the reduction of nonspecific immunosuppression. Currently, Tregs are being tested in clinical trials; however, outstanding questions regarding stability, specificity, and longevity of transferred cells remain. The aim of this research was to better understand the potential plasticity of Tregs, develop novel methods of creating antigen specific Tregs, and determine the optimal signals for Tregs to persist after transfer. To better understand the pathological conversion of Tregs to inflammatory cells, I examined the phenotype of Tregs in systemic sclerosis, a Th2-biased disease. I found that Tregs in patient skin and blood had acquired Th2-cytokine and homing marker expression, respectively, and that both tissue-localized and homing cells express the receptor for IL-33, which was expressed in patient skin. This work suggests that sub-populations of Tregs have the capacity to become pathogenic upon encountering tissue-specific inflammatory signals. Next, in order to create antigen specific Tregs, I developed a novel chimeric antigen receptor (CAR) against HLA-A2 and tested its function. A2CAR-Tregs were highly activated and proliferative in response to HLA-A2, but they retained their suppressive capacity and expression of the transcription factors FOXP3 and Helios, and the effector molecules CD25 and CTLA-4. A2CAR-Tregs were also superior to polyclonal Tregs at preventing xenoGVHD in mice, even at low doses. Thus, A2CAR-Treg cell therapy is a promising new technology to create potent Tregs for antigen-specific transplant tolerance. Finally, to optimize Treg activity and persistence in patients, I developed a series of CARs containing different co-stimulatory domains. Four TNFR superfamily domains were cloned, from 4-1BB, OX40, GITR and TNFR2, and three signalling domains were cloned from the B7/CD28 superfamily, from ICOS, PD-1, and CTLA-4. 4-1BB, OX40, ICOS, and PD-1 containing CARs were expressed on the surface of cells. These tools will provide valuable information in future research on Treg survival in vivo. Collectively, these studies have provided insights to improve both the safety and efficacy of Treg cell therapy.

View record

Because of their potent suppressive capacity and critical role in the normal function of the human immune system, T regulatory cells (Tregs) have long been considered candidates for the therapeutic treatment of autoimmune and chronic inflammatory diseases. However, the clinical implementation of these cells has proven challenging in practice, in part due to a lack of knowledge surrounding this T cell subset. Specifically, an evaluation of the unique functions of individual Treg cell lineages, along with a comprehensive investigation of the non-suppressive capacities of these cells, including chemokine production, is necessary. Furthermore, in the application of Treg cellular therapy in mucosal diseases such as inflammatory bowel disease, the identification of putative antigens that can be targeted by Tregs is warranted. To these aims, I evaluated the phenotypic and functional characteristics of Helios⁺ and Helios⁻ Treg subsets, with the knowledge that expression of Helios, an Ikaros family transcription factor, may differentiate natural, thymic derived Tregs from their in vivo peripherally induced counterparts. I found that Helios positive and negative Treg subsets expressed similar Treg markers and displayed a similar capacity for suppression and plasticity. However, these Tregs did differ in terms of cytokine/chemokine production as well as methylation state of the FOXP3 Treg-specific demethylated region. Futhermore, total populations of FOXP3⁺ Tregs were evaluated for chemokine expression; I found that Tregs produce significant quantities of CXCL8 and other acute phase chemokines, and are able to attract inflammatory cells of the innate immune system. In addition, FOXP3 expression enhances CXCL8 production, likely because of its ability to bind the CXCL8 gene promoter. To evaluate putative antigens that can be targeted by Treg therapy in inflammatory bowel disease, I assessed the role of flagellin in disease. Flagellin exacerbates colitic disease in mice in a TLR5 independent manner and flagellin-specific T cells can be identified in patients with CD. Collectively, these findings bring us closer to the effective application of Treg cellular therapy in the setting of mucosal disease.

View record

FOXP3⁺ T regulatory cells (Tregs) normally function to restrain immune responses, but when their activities go awry diseases such as autoimmunity and cancer can result. Animal models have proven that enhancing or inhibiting the function of Tregs is an effective way to prevent, and in some cases cure, many immune-mediated diseases. Approaches to specifically modulate the activity of Tregs are already being translated to humans, yet we know remarkably little about how Tregs achieve their potent immunosuppressive effects. The aim of this research was to further understand the factors that regulate the molecular phenotype and functionality of Tregs in order to better use them for therapeutic purposes. To achieve this goal, the interaction between Tregs and adenoviral-transduced human monocyte-derived dendritic cells in the context of cancer immunotherapy was explored. I found that these genetically-engineered DCs designed to boost the immune response were still susceptible to Treg suppressive influence. Next, I investigated the biological relevance of chemokine secretion by Tregs and determined that chemokine-mediated active recruitment of their targets of inhibition may be a novel mechanism of action. Finally, I established a human Treg-specific gene signature using Affymetrix microarray technology in order to define better ways to isolate and track these cells. Taken together, these studies have contributed significantly to understanding how Tregs exert their homeostatic control of immunity and revealed potential tactics to manipulate their activity in clinical aspects.

View record

The immune system eliminates threats to the body, but it must also prevent immune-mediated damage caused by inflammation and autoimmune disease. One way that immune responses are limited is by specialized T cells known as T regulatory (Treg) cells. The transcription factor forkhead box P3 (FOXP3) is highly expressed in Treg cells and is critical for their suppressive function. The importance of FOXP3 is demonstrated in humans with a severe autoimmune disease called immune dysregulation, polyendocrinopathy, enteropathy, X-linked (IPEX) caused by mutations in FOXP3. Furthermore, conventional T (Tconv) cells can be re-programmed into suppressive cells upon stable over-expression of FOXP3. Therefore, there is tremendous interest in manipulating FOXP3 function and/or using Treg cells as a cellular therapy to modify immune responses of cancer patients, patients suffering from autoimmune disease, and transplantation patients. To better understand the function of FOXP3, the first goal was to investigate how mutant forms of FOXP3 found in IPEX patients were defective at programming Treg characteristics. Surprisingly, mutant forms of FOXP3 were not completely deficient at converting Tconv cells into Treg cells, suggesting that factors besides a defect in Treg cells may contribute to IPEX pathogenesis. FOXP3 is transiently up-regulated in human Tconv cells upon activation, but its role in these cells is unknown. Hence, the second goal was to examine the function of FOXP3 in Tconv cells by comparing FOXP3-deficient with wild type Tconv cells. FOXP3-deficient Tconv cells proliferated more and produced more cytokines than wild type Tconv cells. This finding suggests that FOXP3 has a role in the regulation of Tconv cell activation, especially in Th17 cells which were found to highly express activation-induced FOXP3. Lastly, the possibility of using FOXP3⁺ cells as a cellular therapy was investigated. A method to expand large, pure populations of human and cynomolgus Treg cells was developed, and ex vivo expanded Treg cells were able to promote mixed chimerism and tolerance to a kidney transplant in cynomolgus macaques. Together, this work sheds light on the role of FOXP3 in CD4⁺ T cell subsets and helps pave the way for use of Treg cell therapy in the clinic.

View record

Adaptive immunity is controlled by CD4+ T cells that develop into functionally distinct T helper (Th) cell subsets. The paradigm that cellular immunity is mediated by Th1 cells and humoral immunity by Th2 cells has been revised upon the discovery of a third Th lineage which produces Interleukin (IL)-17 (Th17 cells). Th17 cells exhibit a unique cytokine profile compared to other Th subsets and are important for host defense against extracellular pathogens. However, aberrant expansion of Th17 cells is linked to several autoimmune diseases including rheumatoid arthritis, psoriasis, inflammatory bowel disease and multiple sclerosis. Due to their association with dysregulated immune responses, there has been intense interest in understanding factors regulating Th17 cell development, phenotype and function. The findings herein identify several mechanisms which regulate this lineage. First, I identified a central role for a transcription factor known as retinoic acid-related orphan receptor C isoform 2 (RORC2) in the development of human Th17 cells. In order to study in vivo derived Th17 cells, a flow cytometry-based sorting method was established to isolate Th17 cells directly from peripheral blood. After defining the phenotype of these bona fide Th17 cells, I investigated their interactions with T regulatory (Treg) cells. Contrary to prior findings and RORC2 overexpression studies, Treg cells were potent inhibitors of in vivo-derived Th17 cell proliferation and suppressed their inflammatory mediators. To examine the stability of this lineage, a comprehensive study of the epigenetic phenotype of Th17 cells was performed. Further to defining Th17 cell’s epigenetic profile, examination of epigenetic marks following exposure to potent polarization conditions suggests Th17 cells are resistant to lineage conversion. Taken together, these findings provide an in-depth examination of this inflammatory T cell subset, provide new tools to study Th17 cells and elucidate mechanisms that regulate Th17 cells, thereby providing insights to aid development of therapeutic strategies for Th17-associated disorders.

View record

The involvement of regulatory T cells (Tregs) in immune homeostasis is now recognized as one of the fundamental mechanisms of immune tolerance. While several different types of Tregs cooperate to establish and maintain immune homeostasis, much current research is focused on defining the characteristics of the CD4⁺CD25⁺ Treg subset, as these cells can mediate dominant, long-lasting and transferable tolerance in many experimental models. The aim of this research was to characterize the biological role of a protein known as forkhead box P3 (FOXP3) that was initially identified as an essential transcription factor for the development of mouse CD4⁺CD25⁺ Tregs, in human CD4⁺ T cells. Following confirmation that, like mouse Tregs, human Tregs also expressed high levels of FOXP3, several approaches were used to investigate the role of this protein in human CD4⁺ T cells. 1) Characterization of endogenous FOXP3 expression in CD4⁺ T cell subsets revealed that this protein is not a Treg-specific marker as was previously thought. Instead, low-level and transient expression was found to be typical of highly activated non-regulatory effector T cells. 2) To generate large numbers of Tregs suitable for cellular therapy, the capacity of ectopic FOXP3 expression to drive Treg generation in vitro was explored. It was found that high and constitutive expression mediated by a lentiviral vector, but not fluctuating expression driven by a retroviral vector, was sufficient to generate suppressive cells. Over-expression strategies were also used to characterize a novel splice isoform unique to human cells, FOXP3Δ2 (FOXP3b). 3) To further probe the requirements of FOXP3 to induce suppressor function, a system for conditionally-active FOXP3 ectopic expression was developed. These studies established that FOXP3 acts a quantitative regulator rather than a “master switch” for Tregs, and that there is a temporal component to its capacity to direct Treg phenotype and function. In summary, this research has significantly expanded the understanding of the biological function of FOXP3 in human CD4⁺ T cells. Based on the potential of these cells to be manipulated for therapy, this work contributes to the field of immunology on both academic and clinical research fronts.

View record