|Home | About | Journals | Submit | Contact Us | Français|
Adoptive transfer in animal models clearly indicate an essential role of CD4+ CD25+ FOXP3+ regulatory T (Treg) cells in prevention and treatment of autoimmune and graft-versus-host disease. Thus, Treg cell therapies and development of drugs that specifically enhance Treg cell function and development represent promising tools to establish dominant tolerance. So far, lack of specific markers to differentiate human Treg cells from activated CD4+ CD25+ effector T cells, which also express FOXP3 at different levels, hampered such an approach. Recent identification of the orphan receptor glycoprotein-A repetitions predominant (GARP or LRRC32) as Treg cell-specific key molecule that dominantly controls FOXP3 via a positive feedback loop opens up new perspectives for molecular and cellular therapies. This brief review focuses on the role of GARP as a safeguard of a complex regulatory network of human Treg cells and its implications for regulatory T cell therapies in autoimmunity and graft-versus-host disease.
Tierexperimentelle Arbeiten zur adoptiven T-Zell-Therapie zeigten eindrücklich die bedeutende Funktion von CD4+ CD25+ FOXP3+ regulatorischen T(Treg)-Zellen in der Prävention und Behandlung von Autoimmunkrankheiten und der Graft-versus-host-Erkrankung. Hierdurch stellt die Anwendung einer adoptiven Treg-Zelltherapie bzw. die Entwicklung von Medikamenten, die die Funktion von Treg-Zellen spezifisch verstärken, eine neue Perspektive zur Etablierung einer dominanten Toleranz dar. Bisher wurde diese Entwicklung durch den Mangel spezifischer Oberflächenmarker zur sicheren Unterscheidung von Treg-Zellen und aktivierten CD4+ CD25+ Effektor-T-Zellen, die ebenfalls FOXP3 in unterschiedlichem Maße exprimieren, gehemmt. Die Entdeckung des Rezeptors «glycoprotein-A repetitions predominant» (GARP bzw. LRRC32) als Treg-Zell-spezifisches Schlüsselmolekül, das dominant FOXP3 in Verbindung mit einem positiven «feedback loop» kontrolliert, eröffnet neue Perspektiven für die zelluläre und molekulare Treg-Zell-Therapie. Diese Übersicht befasst sich mit den Perspektiven, die die Entdeckung von GARP als «safeguard» eines komplexen regulatorischen Netzwerks humaner Treg-Zellen mit sich bringen, sowie mit deren Bedeutung für regulatorische T-Zell-Therapien bei Autoimmunkrankheiten and der Graft-versus-host-Erkrankung.
Negative selection in thymus does represent an imperfect process and thus self-reacting T cell clones do escape central tolerance. Therefore, to keep these potential dangerous self-reactive T cells in check, a multitude of peripheral tolerance mechanisms exist. Dominant tolerance exerted by CD4+ regulatory T (Treg) cells characterized by the expression of CD25 (IL2 receptor alpha)  does represent a major mechanism for immune homeostasis. A major breakthrough in the understanding of this peculiar subpopulation of CD4+ T cells was the identification of the forkhead transcription factor FOXP3 that is essential for their development and function [2,3,4]. The roles of FOXP3 and Treg cells are compellingly demonstrated by the natural loss-of-function, leading to the development of a fatal autoimmune lymphoproliferative disorder in man (IPEX, immunodysregulation, polyendocrinopathy, and enteropathy, X-linked) and mice (scurfy) . Despite advances since the identification of FOXP3 for understanding Treg cell lineage, commitment and function differences between mice and man have become apparent . Contrary to murine Treg cells, which are expressing FOXP3 as lineage-specific marker, consensus has now been reached that T cell receptor (TCR) stimulation of human nonregulatory CD4+ CD25- Th cells does induce FOXP3 expression without interfering with the expression of effector cytokines like IL2 and IFNy or the acquisition of suppressor function [7,8,9,10]. Moreover, even antigen-specific Th lines and clones can express FOXP3 to different extends [11,12,13,14,15]. With that, FOXP3 does not represent a bona fide marker of human Treg cells, suggesting the existence of a Treg cell-specific marker and/or higher-order regulatory networks to explain the qualitative and quantitative differences of FOXP3 expression of human Treg compared to conventional Th cells [16, 17].
The question of Treg cell-specific control mechanisms has been answered only recently with the identification of the orphan receptor glycoprotein-A repetitions predominant (GARP or LRRC32). GARP does represent a Treg cell-specific molecule that dominantly controls FOXP3 via a positive feedback loop (see below) [13, 18, 19]. Treg cell-specific expression of GARP has been further confirmed at the DNA level with the identification of a hypomethylated region in intron 1 together with two differentially methylated regions with enhancer function in human Treg cells . Thus, lineage-directing transcription of GARP in Treg cells is based most likely on differential accessibility of the locus similar to what has been described for FOXP3 [15, 21].
The identification of GARP was enabled by differential gene expression analysis of TCR stimulated CD4+ CD25+ Treg cells compared to their CD4+ CD25- Th cell counterparts (www.ncbi.nlm.nih.gov/geo/query/acc.cgi7acc=GSE13017) as TCR stimulation represents the prerequisite to activate their mutually exclusive functions. The functional contribution of GARP to the genetic program and phenotype of Treg cells was elucidated by ectopic overexpression in human alloantigen-specific Th cells and downregulation of GARP in human alloantigen-specific Treg cells. Retroviral overexpression of GARP in Th cells led to an efficient and stable re-programming of effector T towards Treg cells. This was associated with induction of constitutive expression of FOXP3, the ß-galactoside binding protein LGALS3, the cystein-endoprotease LGMN, and an extended Treg cell signature similar to Treg cells . Interestingly, LGALS3, recently identified FOXP3-dependent gene , and LGMN do represent minor constituents of this feedback loop as they enhance GARP and FOXP3 expression following TCR activation. LGALS represents an interesting regulator of anergy as it has been shown to directly interfere with TCR proximal signaling in human T cells . Furthermore, FOXP3 induction via ectopic expression of LGALS3 depends on an intact serine-6 CK1 phosphorylation site , whereas kinase inhibition of CK1 in Treg cells downregulates FOXP3 expression (fig. (fig.1).1). The result of the GARP-dependent re-programming process is established only after some rounds of antigen-specific TCR-(re-)stimulation. Surprisingly, GARP was more efficient, and the regulatory phenotype induced was stable compared to the overexpression of FOXP3. The final outcome of this complex transcriptional re-programming/ transdifferentiation of effector Th cells towards Treg cells was obvious as most of the Treg cell signature genes showed a sustained high level of expression and induction of effector cytokines was lost [13, 18].
Contrary, downregulation of GARP in human Treg cells, achieved by lentiviral transduction of specific small interfering RNA (siRNA), significantly impaired suppressor function and FOXP3 expression. This was associated with impaired induction of CD83 and CD27, both representing known FOXP3-regulating receptors [13, 23, 24]. Strikingly, lentiviral downregulation of FOXP3 with specific siRNA in Treg cells resulted in similar phenotypic changes and affected induction of GARP, CD83, and CD27. Therefore, a GARP-FOXP3-positive feedback loop in human Treg cells and a close interrelation with other FOXP3-regulating cell surface receptor systems was compellingly established. With that, speculations on a higher-order regulatory network and the necessity on a new molecular definition of human Treg cells [16, 17] reached a robust conceptual framework (fig. (fig.2).2). Moreover, as TGFß-induced Treg cells do not upregulate GARP, lack of GARP might explain the only transient and partial phenotype of TGFß1-induced regulatory Th cells , corroborating the need for a safeguard of FOXP3 expression and maintenance of the regulatory program. In sum, GARP is a key receptor controlling FOXP3 in Treg cells following TCR activation in a positive feedback loop, representing a promising new system for the therapeutic manipulation of T cells in human immune diseases (fig. (fig.33).
Because of the encouraging results of adoptive transfer showing an essential role of CD4+ CD25+ FOXP3+ Treg cells in prevention and treatment of autoimmune [25,26,27] and graft-versus-host disease (GvHD) [28,29,30] in animal models, first clinical trials have been started with CD4+ CD25+ T cells in humans [31, 32]. These trials should also provide insights into the potential effects of Treg cell therapies on the graft-versus-leukemia effect, which is linked in part to GvHD . But keeping in mind the difficulties to achieve pure Treg cell preparations with using CD4+ CD25+ as surrogate marker of human Treg cells , more reliable surface markers to exclude potential hazardous Th cell contaminations from Treg cell preparations have to be used. Such markers should be lineage-specific and indicators of proper regulatory function. GARP fulfills these demands as lineage-specific Treg marker that is responsible for maintenance of the regulatory program (fig. (fig.3).3). Concerning the limitations of Treg cell availability and the dominant role of GARP leading to stable FOXP3 expression and regulatory function in terminal differentiated alloantigen-specific Th cells , genetically engineering disease-associated Th cells would represent an attractive alternative (fig. (fig.33).
In order to facilitate the transformation of research-based cell cultures and processes into clinical good manufacturing practice (GMP) procedures which are required for clinical application, a number of hurdles have to be taken. A strict standardization of the production is one of the most important parameters. Minimizing risks of contamination, especially microbial contamination during the cultivation process, is another important prerequisite for clinical application. Therefore, the clinical acceptance will be greatly improved with cell-cultivating systems which are modular closed systems and easy to use.
As we described in the first part of this review, Treg cells represent a critical component of the immune regulatory system. Through availability of specific antibodies and improved understanding of functional and genetic differentiation of Treg cells, we are now able to better define these T cells and standardize their isolation and quality control (fig. (fig.3).3). An important point still represents the limited quantity of Treg cells, representing a rare subset of CD4+ T cells estimated to be 2-4% or even lower, concerning proper gating to sort pure Treg cells [35,36,37]. Thus, ex vivo expansion is necessary but hindered by the fact that Treg cells replicate slowly in comparison to conventional Th cells . In other words, if a batch of Treg cells is contaminated with other Th cells, these will in the end overgrow the Treg cell population, necessitating a re-purification step (fig. (fig.3).3). This problem has been noticed with most surface markers of human Treg cells (table (table1).1). Therefore, combinations of separate markers for primary enrichment of Treg cells ex vivo (e.g. CD4+ CD25+ CD127lo ) and later re-purification using other markers like GARP or other molecules differentially expressed on Treg cells (table (table1)1) seems to be the future concept to get more pure Treg preparations .
In the research situation most of the isolation procedures were performed using FACS sorting, a cell sorting methodology based upon physical properties of the cellular elements and use of fluorescently labelled antibodies against specific cellular membrane receptors , which is time consuming. Thus, Riley et al.  stated the cumbersome FACS isolation of this infrequent cell type by the words ‘take a billion or so and call me in the morning’. Yet there is another important point with respect to contamination. FACS sorting is a relative open system with numerous porte d'entrées for microorganisms. This problem gets even greater concerning the need for several separation steps for enrichment and re-purification of antigen-specific Treg cells for cellular therapies, discussed above. In conclusion it will be difficult to obtain sufficient Treg cells under GMP conditions using this approach.
Modern cytapheresis procedures are the basis for harvesting sufficient numbers of mononuclear cells (1010-1011) within an acceptable amount of time. This is not only possible for healthy donors but can be achieved with patients as well [40,41,42]. The collected mononuclear cells are harvested under the same conditions as classical blood products such as thrombo-cytapheresis products and are suitable for further processing in GMP facilities.
Novel immunomagnetic techniques in combination with specific antibodies enable the rapid isolation of cells in cell culture bags. This approach enables isolation and cultivation of a large amount of Treg cells in a GMP setting and therefore facilitates the clinical applicability of these cells [39, 43]. This approach also includes the possibility of ex vivo generation of human alloantigen-specific Treg cells at the step of in vitro expansion by means of antigen-specific stimulation with the use of a primary enrichment and a second re-purification step including GARP and other potential markers of Treg cells (table (table1)1) . Further protocols enable the ex vivo expansion of antigen-specific Th cells and genetic modifications of these cells [13, 44]. Although attractive, these approaches however are still rarely found in clinical settings.
Other approaches result from the better understanding of the Treg cells and try to modify Treg cells responses pharmacologically, for example by using antibodies, cytokines, or small molecules  (fig. (fig.11).
Combining the knowledge we have on Treg cells with efficient medical devices for harvesting (apheresis) and isolation (e.g. clinimacs) we are gradually coming in a position in which we will see clinical applications starting.
The identification of GARP as lineage-restricted key receptor of human Treg cells and the characterization of a GARP-FOXP3 positive feedback loop as essential component for the maintenance of the regulatory phenotype does provide a conceptual framework for a new molecular definition of the regulatory program. Open questions remaining include the signal transduction pathway of GARP and the potential ligand. Development of potential GARP-ligand mimetics to enhance regulatory functions in vivo should take into account expression and function of GARP in platelets [46, 47]. In-depth elucidation of the GARP-FOXP3 regulatory system does represent a major challenge for the future and opens up the possibility for generation of antigen-specific Treg cells for adoptive cellular therapies (fig. (fig.3)3) and identification of new molecular targets to develop new strategies to enhance Treg cells in autoimmune diseases and transplantation.
The authors declared no conflict of interest.
Michael Probst-Kepper was supported by grants of the VolkswagenStiftung and Deutsche Forschungsgemeinschaft.