Antigen-Specific Suppression and Immunological Synapse Formation by Regulatory T Cells Require the Mst1 Kinase

Although the cell-to-cell contact between CD4+Foxp3+ regulatory T (Treg) and their target cells is important for the suppressor function of Treg cells, the regulation of this process is not well understood. Here we show that the Mst1 kinase plays a critical role in the suppressor function of Treg cells through regulation of cell contact dependent processes. Mst1 -/- Treg cells failed to prevent the development of experimental colitis and antigen-specific suppression of naïve T cells proliferation in vitro. Mst1 -/- Treg cells exhibited defective interactions with antigen-presenting dendritic cells (DCs), resulting in reduced down-regulation of costimulatory molecules. While wild-type CD4+ Foxp3+ Treg cells formed mobile immunological synapses on supported planar membrane, Mst1 -/- Treg cells did not exhibit ICAM-1 ring or central peptide-MHC clustering. Using two-photon imaging we showed that antigen-specific wild-type Treg cells exhibited dynamic mobile contacts with antigen-pulsed DCs bearing stably associated naïve T cells. In contrast, Mst1 -/- Treg had impairments in their interactions with DCs. Thus, Mst1 is required for Treg cells to mediate contact-dependent suppressor functions.


Introduction
Regulatory T (Treg) cells exert suppressor function in T cell responses to self-antigen, microbial pathogens, transplants, and tumors. Treg cell-mediated suppression in the priming and effector phases of T cell responses involves cell-to-cell contactdependent process as well as bystander suppression [1,2]. Treg cells act on antigen-presenting dendritic cells (DCs) by inhibiting their function through down-modulation of costimulatory molecules [3,4] or by inducing perforin-dependent cell death [5]. Intravital two-photon imaging has shown that the absence of Treg cells prolongs contact duration between DCs and T cells specific for self-antigens [6,7], tumor-related antigens [5], and foreign antigens [8]. Thus, Treg cells can inhibit antigen-induced stable contacts between T cells and DCs, thereby suppressing self-reactive T cells and low-avidity T-cell priming. Adoptively transferred in vitro-expanded isletspecific Treg cells were shown to stably interact with DCs in pancreatic lymph nodes (LN), similar to activated T H cells [7]. However, it remains unclear how natural Treg cells interact with antigen-loaded DCs to suppress T cell priming.
An in vitro study showed that antigen-specific natural Treg cells formed conjugates with antigen-loaded DCs more efficiently than naïve T cells with the same specificity, suggesting that Treg cells could outcompete naïve T cells for antigen-loading DCs, thereby suppressing T cell priming [9]. The conjugate of Treg cells and DCs formed via LFA-1/ ICAM-1-dependent adhesion was required for the suppressor function of Treg cells [9]. Indeed, LFA-1 is more highly expressed in Treg cells than naïve T cells and it is required for efficient generation and function of Treg cell in mice [10,11] and in humans [12]. The immunological synapse (IS) refers to molecular organization of the antigen-specific T cell-APC contact area and it is characterized with segregation of central TCR/pMHC (central supramolecular activation cluster, cSMAC) and peripheral LFA-1/ICAM-1 (pSMAC) [13], which can be reproduced on supported planar membrane [14,15]. Proper organization and stability of IS are regulated by several factors. PKCθ localized in cSMAC destabilizes the IS and promotes mobile synapse in naïve T cells [16], whereas PKCθ is sequestered away from the IS of Treg cells and inhibits Treg suppressor function [17]. Talin, a critical regulator of integrindependent adhesion [18], are associated with pSMAC. Talin1 deficiency severely diminishes antigen-specific T-dC interactions through abrogation of LFA-1 and ICAM-1 interaction, but LFA-1 cluster at T-dC contacts remains intact [19]. Although LFA-1 has been well characterized in multiple immune functions including lymphocyte homing, antigendependent T cell-antigen presenting cell (APC) interactions, and cytotoxicity, elucidation of the regulatory mechanisms of LFA-1 in Treg cells is still limited.
We previously showed that Mst1, the mammalian homolog of Drosophila Hpo, interacted with the Rap1 binding protein RAPL and transmitted signals that chemokine-induced LFA-1dependent lymphocyte adhesion and TCR-dependent adhesion and immunological synapse formation with APC [20]. Mst1 -/mice showed defective lymphocyte trafficking and thymocyte egress [21][22][23]. In addition, Mst1 -/-mice exhibited inefficient thymocyte selection and autoimmune-like disorders with age, suggesting that Mst1-regulation of LFA-1 play an important role in the maintenance of self-tolerance [24]. To elucidate the role of Mst1 in Treg cells, we investigated whether Mst1 deficiency affect the suppressor function and adhesive behaviors of Treg cells. We found that Mst1 is required for the suppression of T cell-induced colitis and T cell proliferation in vitro by natural Foxp3 + CD4 + Treg cells. Two-photon imaging within LN tissues and immunological synapse formation on supported planar membrane revealed the dynamic interactions of Treg cells with antigen-loaded DCs with mobile immunological synapse (IS) formation. Mst1 -/-Treg cells are severely impaired in interactions with DCs, which is characterized by the defect of both cSMAC and pSMAC formations, therefore indicating that Mst1 plays a crucial role in shaping Treg cell suppressor functions by regulating their adhesive behaviors.

Mice
Mst1 -/-mice on a C57BL/6 background were previously described [21]. C57BL/6 mice were purchased from CLEA Japan. OT-II mice on a C57BL/6 background [25] were obtained from Jackson Laboratories. Foxp3-IRES-GFP knockin (Foxp3-GFP) mice [26] were kindly provided by Dr. M. Hori (RIKEN, Yokohama, Japan). Mst1 -/-, OT-II, and Foxp3-GFP mice were crossed to generate Mst1 -/-OT-II mice expressing Foxp3-GFP. RAG2 -/-mice [27] were obtained from Center for Animal Resources and Development, Kumamoto University (Kumamoto, Japan). All mice were maintained under specific pathogen-free conditions in the animal facility at Kansai Medical University (Osaka, Japan) and were treated and used for experiments in accordance with institutional guidelines and ethical approvals for the care of experimental animals. The protocols were approved by the Animal Care and Use Committee of Kansai Medical University (#13-015).

Cell preparation
T cells were isolated from spleens and lymph nodes of 12-to 20-wk-old mice by magnetic-activated cell sorting using a pan-T cell isolation kit II (Miltenyi Biotec, Auburn, CA). Isolated T cells were stained with APC-labeled anti-CD8, PerCP-Cy5.5 labeled anti-CD62L antibodies. CD62L hi CD8 -Foxp3-GFPor CD8 -Foxp3-GFP + cells were sorted as naïve T or Treg cells, respectively, by using a FACS Aria cell sorter (BD Bioscience). The purity of naïve T and Treg cells were greater than 97%. For preparation of CD11c + dendritic cells, spleens of C57BL/6 mice were digested with collagenase D (1 mg/ml) and DNase I (50µg/ml) for 30 min at 37°C and CD11c + cells were isolated by magnetic-activated cell sorting using anti-CD11c beads (Miltenyi Biotec). Dendritic cells were generated in vitro from bone marrow cells of C57BL/6 mice with optimal concentration of GM-CSF-conditioned medium as described [28].

Immunological synapse assay
Mouse ICAM-1-GPI was purified, fluorescently labeled with Cy3, and incorporated into liposomes consisting of 0.4 mM DOPC containing 0.1 mol% biotin-CAP-PE (Avanti Polar Lipids, Alabaster, AL) essentially as described [30]. ICAM-1 liposomes were mixed with DOPC/CAP-biotin-PE liposomes containing 1 mol% DGS-NTA-Ni at a 2:1 ratio and deposited onto piranhatreated cover glass mounted on a 35 mm dish. Lipid bilayers on the glass were blocked with 1% BSA and loaded sequentially with Alexa488-labeled streptavidin (1 µg/ml) and monobiotinylated I-A b with 0.1 M OVA peptide (AAHAEINEA) obtained through the NIH Tetramer Core Facility (Atlanta, GA) and CD80-histidine tag (Creative Biomart, Shirley, NY). Densities of ICAM-1 (250 molecules/μm 2 ) and CD80 (90 molecules/μm 2 ) were measured as described [30], and supported pMHC-dependent attachment and IS formation on lipid bilayers. Treg cells (1×10 5 ) in phenol-red free RPMI-1640/5%FCS were added to lipid bilayers, and live-cell imaging was performed in a humidified 5% CO 2 incubator chamber on the stage of an inverted Olympus IX81 microscope (Olympus, Tokyo, Japan) equipped with a Hamamatsu C9100-13 EM-CCD camera and 60× objective lens (NA 1.45). Time-lapse images were acquired for peptide-MHC in the epiillumination mode and for ICAM-1 in the TIRF mode with a 564 nm solid-state laser (Andor, Belfast, UK). All data were collected and analyzed using Metamorph software (Molecular Devices, Sunnyvale, CA).

Two-photon imaging using LN slices
CD11c + BMDCs (1×10 6 ) were stimulated with 1 µg/ml lipopolysaccharide for 16 h and labeled with CFSE prior to loading with or without 1 µM OVA 323-339 , and were subcutaneously injected into the front paw of C57BL/6 mice. After 24 h, brachial LNs were isolated, embedded in 4% lowmelting agarose gels, and then cut open with a vibratome (DSK, Kyoto, Japan). OT-II naïve T cells, Mst1 +/+ or Mst1 -/-OT-II Treg cells were labeled with CMAC or CMTMR. The mixture of these cells was applied to the cut LN surface. After incubation for 30 min at 37°C/5% CO 2 , the LN slice was mounted on a RC-26 flow chamber (Warner Instruments, Hamden, CT), with a supply of warm RPMI1640 medium saturated with a 95% O 2 /5% CO 2 gas. Two-photon imaging was set up using a FV1000 upright microscope (Olympus), with a Ti: sapphire laser (MaiTai HP DeepSee-OL, Spectra-Physics, Irvine, CA) and a 25X/1.05 NA objective (XLPLN25XWMP, Olympus). The images were taken with excitation wavelength at 820 nm. Emitted light was passed through 420-460 nm, 495-540 nm, and 575-630 nm bandpass filters to detect CMAC, CFSE, and CMTMR fluorescence, respectively. Timelapse x-y images with 2.5 µm apart of z-stacks were taken every 20 sec for 20 min. Volocity software (PerkinElmer, Waltham, MA) was used for 3D cell tracking and calculation of cell motility parameters.

Defective suppressor functions of Mst1 -/-Treg cells
In Mst1 -/-mice, Foxp3 + CD4 + Treg cells in the thymus were decreased but recovered in the periphery [24] and exhibited comparable levels of Foxp3 and surface receptors including CTLA-4 ( Figure S1). To determine whether the Mst1 deficiency altered Foxp3 + Treg cell functions, we employed the experimental colitis model induced by adoptive transfer of naïve T cells into Rag2 -/-mice. In this model, the co-transfer of Treg cells into lymphopenic mice prevents the colitis [31]. Indeed, defective weight gain ( Figure 1A) and development of the colitis ( Figure 1B) were induced by adoptive transfer of CD62L hi naïve T cells, but these symptoms were attenuated by co-transfer of Foxp3 + Treg cells. In contrast, Mst1 -/-Foxp3 + Treg cells were not able to prevent colitis and showed a similar degree of the body weight loss as mice that received only naïve T cells. Histological examination confirmed that the mice that received naïve T cells and Mst1 -/-Treg cells exhibited intestinal inflammation with loss of crypts and infiltration of leukocytes, resulting in intestinal thickening and shortening ( Figure 1A, 1B).
Since Mst1 is required for lymphocyte trafficking, the diminished ability of Mst1 -/-Treg cells to suppress experimental colitis may be due to the impaired trafficking of Mst1 -/-Treg cells. To confirm defective suppressor functions of Mst1 -/-Treg cells directly, we examined in vitro suppression of naïve T cell proliferation. Both wild-type and Mst1 -/-Treg cells suppressed naïve T cell proliferation comparably in response to antigenindependent stimulation using anti-CD3 antibody (Figure 2A). To evaluate antigen-specific suppression, we prepared Treg cells and naïve T cells from Mst1 +/+ OT-II or Mst1 -/-OT-II mice. Compared to Mst1 +/+ OT-II Treg cells, Mst1 -/-Treg cells were less efficient at mediating the suppression of naïve OT-II CD4 + T cell proliferation upon stimulation with OVA peptide and splenic DCs (Figure 2B). Since Mst1-deficiency did not influence on anti-CD3-induced suppressor function, Mst1 is dispensable for the downstream events of TCR stimulation in Treg cells, instead it is more likely required for the contacts between Treg cells and antigen-loaded DCs.

Impairments of antigen-dependent conjugate formation of Mst1 -/-Treg cells
To examine this possibility, we co-cultured OT-II Foxp3 + CD4 + Treg cells and OT-II CD62L hi Foxp3 -CD4 + naïve T cells with CD11c + splenic DCs. When mixed with an equal ratio of wild-type Treg and naive T cells, they formed conjugates with the OVA peptide-pulsed DCs ( Figure 3A). This antigen-specific conjugate formation was inhibited almost completely with antibodies to LFA-1 and ICAM-1 (not depicted), as reported [9].
Wild-type Treg cells formed conjugates with DCs more efficiently than naïve T cells; the relative number of Treg cells in each cluster ranged from 0.75 to 3.33 with a median of 1.33, which is in agreement with the previous study [9]. On the other hand, Mst1 -/-OT-II Treg cells adhered poorly to DCs compared to naïve T cells in the same condition, resulting in an average ratio of Mst1 -/-Treg cells and naïve T cells in each cluster ranging from 0.14 to 2.5 with a median of 0.80. To directly compare the ability of Mst1 +/+ and Mst1 -/-Treg cells to form antigen-dependent conjugates, equal numbers of Mst1 +/+ or Mst1 -/-Treg cells were mixed with peptide-pulsed DCs. Mst1 -/-Treg cells formed conjugates with DCs at a 60% lower rate than Mst1 +/+ Treg cells ( Figure 3B). Thus, Mst1 is required for efficient antigen-specific contacts with DCs through LFA-1.
To examine the consequence of this defect in Mst1 -/-Treg cells on contact-dependent suppressor functions, we measured the down-regulation of co-stimulatory molecules in DCs [3,4,9]. Co-culture of OT-II naïve T cells and peptide-pulsed DCs increased CD86 expression by DCs, which was almost completely abrogated by the addition of wild-type OT-II Treg cells ( Figure 3C). However, Mst1 -/-Treg cells partially inhibited the CD86 up-regulation only by a half, compared to wild-type Treg cells (Figure 3C). This result supported the notion that inefficient conjugate formation between Mst1 -/-Treg cells and DCs impairs the ability of Treg cells to mediate suppressor functions.

Defective immunological synapse formation of Mst1 -/-Treg cells
To clarify the impairment of antigen-specific contacts of Mst1 -/-Treg cells with DCs, we examined IS formation using a planar lipid membrane expressing OVA peptide/I-A b , ICAM-1, and CD80 molecules. Foxp3 + OT-II Treg cells showed rapid formation of IS within a couple of minutes, characterized by a central pMHC cluster surrounding ICAM-1 and this was indistinguishable from that of naïve T cells (Figure 4A, Video S1). The absence of the OVA peptide/MHC did not induce the IS formation (not depicted). Compared to naïve T cells, we found that Treg synapses were motile; symmetrical IS frequently broke by the partial loss of pSMAC with development of the leading edge and uropod (Figure 4B, Video  S2). Cell tracking analysis showed that naïve T cells were basically sessile, while Treg cells frequently migrated on the planar lipid bilayers (Figure 4C). On the other hand, Mst1 -/-Treg cells exhibited inefficient attachment ( Figure 2S) and poorly developed mature IS (Figure 4A, 4B, Video S3). Although Mst1 -/-Treg cells exhibited some ICAM-1 clusters, they did not distribute in a ring structure and pMHC clusters were absent, even with increased periods of incubation ( Figure  4A, 4B). Thus, these results clearly indicate that Mst1 is required for efficient attachment via LFA-1/ICAM-1 and development of pSMAC and cSMAC formation.

Interactions of natural Treg cells with DCs in lymphoid tissues
Since Mst1 also regulates interstitial migration [21], the combined effects of impaired motility and IS formation may aggravate in vivo Treg suppression, as seen in the experimental colitis model (Figure 1). To clarify this point, Treg cell trafficking and interactions with DCs should be examined. However, two-photon imaging of LNs with conventional intravital or explant methods after adoptively transfer of Treg cells was rather difficult due to the limited number of natural Foxp3 + Treg cells or induced Treg cells available from Mst1 -/mice (not depicted). To circumvent this, we took advantage of the LN slice method, which allowed us to apply relatively small numbers of Treg cells together with naïve T cells directly to the parenchyma of LNs that had been transferred with antigenloaded DCs. Under this protocol we performed two-photon imaging to examine the interstitial migration and interaction of Treg cells with DCs within LN tissues. Foxp3 + OT-II Treg cells and naïve OT-II T cells were applied to the sliced tissues of peripheral LNs in which OVA-pulsed DCs had been transferred subcutaneously. Without OVA antigen, OT-II Treg and naïve T cells exhibited comparable motility with approximately 10 µm/min average velocity (Figure 5A). In the presence of OVA antigen, naïve T cells arrested and formed stable contacts with DCs ( Figure 5B, 5C, Video S4). As a result, average velocity was as low as 5 µm/min ( Figure 5A). Treg cells exhibited transient arrest and swarmed around DCs bearing naïve T cells (Figure 5B, 5C, Video S4); the contact duration of Treg cells was shorter than that of naïve T cells, resulting in intermediate velocities ( Figure 5A). Representative cell migration paths showed that naïve T cells were almost sessile and wild-type Treg cells exhibited a brief "stop and go" pattern ( Figure 5B). In contrast, Mst1 -/-Treg cells migrated at intermediate velocities (7.6 µm/min) without the OVA antigen, indicating that Mst1 deficiency cause inefficient Treg migration ( Figure 5A). Furthermore, Mst1 -/-Treg cells did not exhibit arrest even in the presence of antigen; Mst1 -/-Treg cells made contacts with DCs, but were easily dissociated and apart from them ( Figure 5B, 5D, 5E, Video S5-6). As a result, there was no significant difference in velocity with or without antigen ( Figure 5A). Thus, the Mst1 deficiency had detrimental effects on both motility and antigen-specific arrest within lymphoid tissues. These combined defects likely explain the severe impairments of Mst1 -/-Treg cells in the suppression of experimental colitis.

Discussion
Here we showed that Mst1 -/-Treg cells failed to prevent the development of experimental colitis and antigen-specific suppression of naïve T cells proliferation in vitro. Mst1 -/-Treg cells exhibited defective interactions with DCs, resulting in impaired down-regulation of costimulatory molecules. While wild-type CD4 + Foxp3 + Treg cells formed mobile immunological synapses on supported planar membrane, Mst1 -/-Treg cells exhibited reduced attachment and did not elicit pSMAC and cSMAC formations. Consistently, antigen-specific wild-type Treg cells exhibited dynamic mobile contacts with antigenpulsed DCs bearing stably associated naïve T cells. In contrast, Mst1 -/-Treg cells were defective in their interactions with DCs. Thus, Mst1 is required for Treg cells to mediate contact-dependent suppressor functions. Since Mst1 is required for medullary migration and self-antigen recognition of mature thymocytes by regulating LFA-1/ICAM-1 interaction [24], LFA-1 regulation seems to be a basic regulatory mechanism by Mst1 that controls adhesion and migration of Treg cells.
It should be noted that natural Treg cells and naïve T cells within LN tissues exhibit a distinct interaction pattern of short- lived contacts and stable contacts with DCs, respectively. This difference is corroborated by mobile IS of Treg cells on a supported planar membrane. The short-lived mobile contacts did not show preferential clustering of Treg cells with DCs in the LN tissues, which was in contrast to the in vitro situations that lacked motility-support environments. Taken together, these observations indicate that short-lived mobile behavior is characteristic of natural Treg cells within LN environments. Considering this aspect of Treg cell interactions with DCs, it is unlikely that Treg cells inhibit naïve T-cell priming by occupying DCs in LNs, but rather via down-modulation of co-stimulatory properties or other short contact-dependent suppression such as perforin-dependent killing of DCs mechanisms [5].
Our results show a novel role of Mst1 in the process of IS formation by natural Treg cells. We previously showed that Mst1 and RAPL were localized in vesicle compartments and relocated to the T-APC contact sites with co-localization with LFA-1 [20,32]. The current study extends the previous studies and indicate that a linkage of spatial organization of LFA-1/ ICAM-1 and TCR/pMHC bindings. Talin1 deficient T cells exhibits unstable T-dC interactions with impaired polarization of vinculin and F-actin at the IS. However, LFA-1 is still recruited to the contact sites with DC [19]. Therefore, talin stabilizes the IS formation through a linkage of LFA-1 to F-actin, whereas recruitment of LFA-1at the contact site is a talin-independent process that requires Mst1.
The mechanism by which Mst1 controls cSMAC formation is not clear. Since the IS formation can be correlated with topologically driven receptor segregation based on the size of the different receptor -ligand pairs [33], impaired large LFA-1/  ICAM-1 cluster formation could also impair small TCR/peptide-MHC segregation. TCR microclusters formed in the periphery of SMAC are centripetally transported via the processes requiring dynein/microtubule [34] and myosin IIA/F-actin [35], or transported toward actin-depolymerization sites [36]. Mst1 deficiency could affect these processes. The recent study reported that Dock8 was one of the Mst1 target upon chemokine stimulation, and double deficiency of Mst1 and Mst2 impaired chemokine-stimulated F-actin development with reduced Rac and Rho activities [23]. Dock8 -/-CD8 T were reported to be defective in ICAM-1 accumulation at the contact site with antigen-loaded DCs [37], which suggest that Dock8 is involved in IS formation. Dock8 is a Cdc42 guanine nucleotide exchange factor, which regulates actin polymerization at the leading edge of DCs and plays a crucial role in interstitial migration of DCs [38]. Thus, Mst1 signals could control cytoskeletal signals involved in cSMAC formation via Dock8 activation. Further studies are needed to clarify the role of Dock8 in Mst1-regulation of LFA-1 in Treg cells.
Treg cells dynamically interact with antigen-loaded DC in vitro [9]. Our results demonstrate dynamic adhesive behavior of antigen-specific Foxp3 + natural Treg cells with antigen-loaded DC within LN tissues, which is distinct from stable adhesion of naïve T cells with DC. The stable OT-II naïve T cell IS within 30 min observation time is in agreement with those of naïve AND or OT-I T cells, which showed stable IS or IS with limited motilities (1.5 µm/min) [16], followed by exhibiting mobile IS at later time points (90min or later) [36]. The OT-II Treg IS formation on the supported planar membrane was indistinguishable from naïve OT-II T cells at the early step, however was quickly transformed into the mobile IS. Notably, Zanin-Zhorov A et al. has recently reported that human CD4 + CD25 + Treg IS exhibited relatively more stable on the supported planar bilayers presenting anti-CD3 mAb and ICAM-1, compared to effector T cells [17]. Although the comparison between human CD4 + CD25 + Treg cells and mouse Foxp3 + Treg cells under the same condition is not available, the disparity to their study might reflect differences in TCR stimulation, Treg subsets, or distinct regulation of mouse and human T cells.
Previous studies showed that the presence of Treg cells inhibited stable contact between antigen-specific naïve T cells and DCs within LN tissues [5][6][7][8]. However, in our experimental conditions of 20-30 min observation time, naïve T cells arrested normally and formed conjugates with DCs in the presence of Treg cells. This could be due to the longer period of time required by Treg cells to down-modulate co-stimulatory molecules on DCs and/or the ability of the high-affinity OT-II TCR override the inhibitory effect of Treg cells on the stable contacts. Indeed, low-affinity, but not high-affinity interactions of OT-I T cells with DCs was affected by antigen-specific Treg cells [8].
CD4 + Foxp3 + Treg cells were reduced in the thymus but recovered in the periphery in Mst1 -/-mice [24]. We noticed that the in vitro induction of Foxp3 + Treg cells from CD4 + naïve T cells with TGFβ was less efficient compared to wild type. However, the levels of Foxp3 and CTLA-4 of induced Mst1 -/-Treg were comparable with intact suppressor functions on T cell proliferation induced by TCR stimulation. The inefficient induction of Treg cells from Mst1 -/-CD4 + naïve T cells suggest that Mst1-regulation of homotypic cellular interactions through LFA-1/ICAMs is involved in the Treg induction processes in vitro, which could be compensated in vivo.
Mst1-deficient mice developed lymphocyte and plasma cell infiltrations into multiple organs with hypergammaglobulinemia and autoantibody productions with age. Combined defects of Treg cells and thymic selections [24] likely underlie impaired self-tolerance in Mst1-deficient mice. Our study sheds a new insight into autoimmunity-like phenotypes with T cell immunodeficiency in Mst1-deficient human [39,40], and could aid in the development of Treg manipulation and therapeutics.