Multipotent adult progenitor cells induce regulatory T cells and promote their suppressive phenotype via TGFβ and monocyte-dependent mechanisms

Dysregulation of the immune system can initiate chronic inflammatory responses that exacerbate disease pathology. Multipotent adult progenitor cells (MAPC cells), an adult adherent bone-marrow derived stromal cell, have been observed to promote the resolution of uncontrolled inflammatory responses in a variety of clinical conditions including acute ischemic stroke, acute myocardial infarction (AMI), graft vs host disease (GvHD), and acute respiratory distress syndrome (ARDS). One of the proposed mechanisms by which MAPC cells modulate immune responses is via the induction of regulatory T cells (Tregs), however, the mechanism(s) involved remains to be fully elucidated. Herein, we demonstrate that, in an in vitro setting, MAPC cells increase Treg frequencies by promoting Treg proliferation and CD4+ T cell differentiation into Tregs. Moreover, MAPC cell-induced Tregs (miTregs) have a more suppressive phenotype characterized by increased expression of CTLA-4, HLA-DR, and PD-L1 and T cell suppression capacity. MAPC cells also promoted Treg activation by inducing CD45RA+ CD45RO+ transitional Tregs. Additionally, we identify transforming growth factor beta (TGFβ) as an essential factor for Treg induction secreted by MAPC cells. Furthermore, inhibition of indoleamine 2, 3-dioxygenase (IDO) resulted in decreased Treg induction by MAPC cells demonstrating IDO involvement. Our studies also show that CD14+ monocytes play a critical role in Treg induction by MAPC cells. Our study describes MAPC cell dependent Treg phenotypic changes and provides evidence of potential mechanisms by which MAPC cells promote Treg differentiation.

Tregs are indispensable players of immune regulation by maintaining self-tolerance and homeostasis. Tregs are characterized as natural Tregs (nTregs), which are developed in the thymus during embryonic state, or induced Tregs (iTregs) that arise from effector T cells in the periphery, preferentially during inflammatory conditions 1 . Tregs express both CD4 and CD25 surface antigens as well as the transcription factor, FoxP3, a critical gene involved in Treg development and function 2,3 . Treg deficient mice suffer fatal autoimmunity called "scurfy mice" 4 . Similarly, humans born with dysfunctional FoxP3 develop an autoimmune syndrome called immunodysregulation polyendocrinopathy enteropathy X-linked (IPEX), which is characterized by severe enteropathy, endocrinopathy, and eczematous dermatitis 5,6 . Tregs control inflammation and modulate the immune system by several mechanisms which can be categorized as: (1) secretion of anti-inflammatory factors such as interleukin 10 (IL-10), interleukin 35 (IL- 35), and TGFβ; (2) metabolic disruption by cyclic adenosine monophosphate (cAMP), CD39, and CD73; (3) inhibition of antigen presenting cell maturation; and (4) induction of effector T cell death by interleukin 2 (IL-2) consumption and granzyme and perforin cytolysis 7,8 . Furthermore, Tregs express co-inhibitory receptors, including cytotoxic T lymphocyte associated protein 4 (CTLA-4) and program cell death protein 1 ligand (PD-L1), that further support Treg immune regulatory properties. Due to its ability to reduce inflammation and modulate the immune system, the use of Tregs as therapy to treat autoimmune diseases is currently being explored 9,10 . MAPC cells are adult bone marrow derived adherent cells with immunomodulatory properties that reduce inflammation by regulation of immune system functions 11 . MAPC cells inhibit allogeneic T-cells in a mixed lymphocyte reaction (MLR) and suppress an allogeneic reaction between two mismatched lymphocyte populations 12,13 . MAPC cells inhibit allogeneic cell and memory response mediated T-cell proliferation in vitro in a dose-dependent manner 14 . In vitro and in vivo studies have also demonstrated that MAPC cells suppress T cell homeostatic expansion driven by IL-7 via prostaglandin E2 (PGE-2) 15,16 . In rat models of traumatic brain injury or stroke, MAPC cell treatment significantly increases Treg frequencies in the spleen and blood [17][18][19] . Concurrently, MAPC cell treatment reduces proliferation of both CD4 + and CD8 + T effector cells 17 . MAPC cell administration also leads to a reduction of pro-inflammatory cytokines in a sheep model of hypoxic ischemia and rat models of stroke or traumatic brain injury [17][18][19][20] .
MAPC cells share many immunomodulatory functions with mesenchymal stem cells (MSCs), however, they are phenotypically and functionally distinct cell types 21 . MAPC cells can be differentiated from MSCs by size and morphological features, differential expression of surface markers such as CD140a and CD140b, and their production of CXCL5 [21][22][23] . Additionally, MAPC cells are cultured in hypoxic conditions and can be expanded to higher population doublings than MSCs making large scale manufacturing more feasible. In a clinical study evaluating the administration of MultiStem ® , a clinical grade product of MAPC cells, in patients receiving a liver transplant, a transient upregulation of Tregs in the blood was observed 24 . MultiStem is currently under clinical evaluation to treat acute ischemic stroke (NCT03545607), ARDS (NCT02611609), and trauma (NCT04533464). In the case of ischemic stroke, a Phase 2 clinical trial revealed that MultiStem treatment not only improved clinical outcomes but also reduced inflammation characterized by decreasing T effector cells and pro-inflammatory cytokines levels in the blood 25 .
While MAPC cell-dependent induction of Tregs has been observed 17,18,24,26 , the mechanism(s) by which MAPC cells increase Tregs remain to be fully elucidated. In this study, we investigated MAPC cell induction of Tregs in vitro and examined the effects of MAPC cells on Treg proliferation, characterization of their suppressive phenotype, and analysis of their expression of CD45 isoforms. We also evaluated the involvement of factors secreted by MAPC cells in the induction of Tregs. These results provide greater insight into the mechanistic pathways in which MAPC cells may modulate inflammation and immune responses in the setting of acute inflammatory diseases including ischemic stroke, ARDS, GvHD and AMI.
T cell stimulation [27][28][29] . The percentage of Tregs (CD3 + CD4 + CD25 + FoxP3 + , see Supplemental Fig. 1 for gating strategy) was determined by flow cytometry at day 7 ( Fig. 1). As seen in vivo, allogeneic MAPC cells increased the percentage of Tregs by approximately 2.7-fold at 2:1 and 2.5-fold at 4:1 PBMC:MAPC cell ratios, respectively (Fig. 1A). MAPC cell induction of Tregs was also assessed at day 4, 5, and 14. At day 4, co-culture of PBMCS with MAPC cells at 2:1 PBMC:MAPC cell ratio increased Tregs frequencies approximately 80% while no Treg induction was observed at 4:1 (Supplemental Fig. 2A). At day 5, MAPC cells also increased Treg frequencies in both 2:1 and 4:1 PBMC:MAPC cell ratios, however, maximum Treg induction was observed at day 7. Augmented cell death was observed after 14-day co-culture (data not shown). Given that in vivo studies have demonstrated that MAPC cells are undetected approximately 7 days after administration 20,30,31 and maximal MAPC cell mediated Treg induction was observed at day 7, induction of Tregs by MAPC cells was studied at day 7. Increased Treg counts were also observed after co-culture with MAPC cells at both day 5 and day 7 (Supplemental Fig. 2B). Further dilutions of MAPC cell concentration demonstrated that MAPC cell induction of Tregs is dose dependent (Fig. 1B). Increased FoxP3 and CD25 expression on Tregs have been shown to correlate with increased anti-inflammatory capacity [32][33][34] . Thus, the FoxP3 and CD25 median fluorescent intensity (MFI) within the Treg population was examined (Fig. 1C,D). Interestingly, miTregs displayed increased FoxP3 and CD25 expression than Tregs from PBMCs cultured alone. Comparable to Tregs, miTregs were Helios positive (~ 80%)   Fig. 2D,E).
To study whether MAPC cells can drive the differentiation of CD4 + T cells into Tregs, CD25 + cells were depleted using magnetic bead separation and CD25 negative PBMCs were co-cultured with MAPC cells. We found that MAPC cells increased the percent of Tregs by 2.6-fold, suggesting that MAPC cells can induce the conversion of CD4 + T cells into Tregs (Supplemental Fig. 3B). Collectively, these data demonstrate that MAPC cell interactions with PBMCs result in the induction of Tregs.

MAPC cells augment Tregs proliferation in vitro.
To examine if MAPC cells induce Treg expansion , Treg proliferation was determined by assessing the expression of Ki67 as a surrogate marker of proliferation in the presence or absence of MAPC cells (Fig. 1E). Indeed, miTregs showed elevated Ki67 expression than Tregs from PBMCs cultured alone. Increased Ki67 expression was only observed at day 7 post co-culture (Supplemental Fig. 2C). To confirmed that MAPC cells indeed induce Treg proliferation, PBMCs were labeled with a cell proliferation dye (cell proliferation dye efluor 450) and after 7 days in culture, Treg proliferation was assessed ( Supplementary Fig. 2F). Dilution of the cell proliferation dye was only observed on Tregs from PBMCs co-cultured with MAPC cells, confirming that MAPC cells indeed induce Treg proliferation. MAPC cells also increased Ki67 expression on Tregs when co-cultured with CD25 negative PBMCs (Supplemental Fig. 3C). Ki67 induction by MAPC cells was restricted to the FoxP3 + Tregs and not FoxP3 − CD4 + T cells (Fig. 1F). Together, these data suggest that MAPC cells preferentially promote the expansion of Tregs but not FoxP3 − CD4 + T cells, as previously reported 14 .
Analysis of PBMC:MAPC cell co-culture supernatant revealed that IL-2 levels were increased when compared to supernatants from PBMCs cultured alone (Supplemental Fig. 2G). Low IL-2 levels were detected in supernatants from PBMCs and MAPC cells cultured alone. To identify the source of IL-2, cell cultures were treated with Golgistop for 18 h and IL-2 was measured intracellularly by flow cytometry. IL-2 was primarily secreted by Tregs (Supplemental Fig. 2H). Conversely, IL-2 was not detected on FoxP3 − CD4 + T cells, MAPC cells, or monocytes (data not shown). These data suggest that MAPC cells induce Treg activation, thereby promoting their secretion of IL-2 which can potentially contribute to Treg proliferation.
Increased suppressive phenotype and function in miTregs. Increased expression of CTLA-4, HLA-DR, and PD-L1 on Tregs has been shown to correlate with a potent suppressive phenotype [35][36][37][38] . The expression of these markers on miTregs was assessed by flow cytometry and compared to Tregs from PBMCs (Fig. 2). miTregs had an increased expression of CTLA-4 ( Fig. 2A), HLA-DR (Fig. 2B), and PD-L1 (Fig. 2C) compared to PBMC Tregs. MAPC cell induction of CTLA-4, HLA-DR, and PD-L1 expression was restricted to Tregs and not FoxP3 − CD4 + T cells (Supplemental Fig. 4). HLA-DR expression was also increased by MAPC cells on Tregs when co-cultured with CD25 negative PBMCs (Supplemental Fig. 2D). To investigate the effects of MAPC cells on Treg suppressive function, a T cell suppression assay was performed (Fig. 2D). miTregs suppressed T cell proliferation more efficiently than PBMC derived Tregs in all Treg: PBMC ratios tested, correlating with an increased suppressive phenotype. In addition to suppressing T cell proliferation more efficiently, miTregs also produced increased levels of TGFβ, also correlating with increased suppressive capacity (Supplemental Fig. 2I). Together, these data demonstrate that MAPC cells not only increase the frequency of Tregs and their expansion but also its activation and suppressive status.

MAPC cell induction of Tregs is TGFβ and IDO dependent. When PBMCs and MAPC cells were co-
cultured in Transwells ® , Treg induction was equivalent to the levels observed when cells were in direct contact, suggesting that MAPC cells promotes Tregs via soluble factors ( Supplementary Fig. 3F). TGFβ has been implicated as a strong Treg inducer both in vivo and in vitro [45][46][47][48] . First, TGFβ and latent associated protein (LAP) levels within culture supernatants were assessed (Fig. 4A,B). Both TGFβ and LAP levels were elevated within the culture supernatants in the presence of MAPC cells when compared to PBMCs alone. Furthermore, TGFβ and LAP levels were also assessed in supernatants from MAPC cells cultured alone at equal MAPC cell numbers as 2:1 PBMC:MAPC cell ratio. The levels of TGFβ and LAP were equivalent to those seen in the 2:1 PBMC:MAPC cell supernatants, suggesting that MAPC cells are the main source of TGFβ. MAPC cell production of TGFβ was confirmed by intracellular staining of both MAPC cells cultured alone or with PBMCs at 2:1 PBMC:MAPC cell ratio (Fig. 4C). TGFβ production is tightly regulated at a posttranscriptional level 49 . Conversion of latent into mature TGFβ requires cleavage from LAP. Latent TGFβ binds to GARP, a type I transmembrane cell surface docking receptor 50,51 . Since LAP and GARP are associated with TGFβ maturation, the expression of LAP and GARP was assessed on MAPC cells (Fig. 4D). Approximately 80% MAPC cells were positive for LAP, whereas 65% of MAPC cells were GARP positive. Given that supernatants of MAPC cells cultured alone have equivalent levels of TGFβ and LAP as the 2:1 PBMC:MAPC cell co-cultures and that MAPC cells have elevated expression TGFβ signaling blockade increased HLA-DR expression on PBMC Tregs, while the expression of CTLA-4 and PD-L1 remained unaltered. This could be due to overall increased HLA-DR expression on FoxP3 − CD4 + cells in the inhibitor treated conditions (Supplemental Fig. 6).
In vitro, MAPC cells inhibit T cell proliferation via the secretion of IDO 14 . Furthermore, IDO has been implicated as an important factor of Treg induction 52,53 . IDO production by MAPC cells was confirmed by intracellular staining of MAPC cells cultured alone or with PBMCs at 2:1 PBMC:MAPC cell ratio (Supplemental Fig. 7A). Analysis of IDO MFI demonstrates that MAPC cells produce similar levels of IDO when cultured alone or with PBMCs. To determine if IDO is involved in MAPC cell induction of Tregs, PBMC and MAPC cells were co-culture in the presence of INCB024360, a selective IDO1 inhibitor. IDO inhibition resulted in a partial reduction of Treg induction by MAPC cells, suggesting that IDO is involved in MAPC cell induction of Tregs, but it is not the primary mechanism (Supplemental Fig. 7B). While Ki67 was still induced in miTregs in the presence of INCB024360, it was reduced when compared to the vehicle treated control (Supplemental Fig. 7C). CTLA-4 induction was not observed in the presence of the IDO inhibitor (Supplemental Fig. 7D). In the presence of IDO blockade, HLA-DR and PD-L1 expression on miTregs were also upregulated. However, HLA-DR was mildly reduced in 2:1 PBMC:MAPC cell condition (Supplemental Fig. 7E), whereas no effect in PD-L1 expression was observed (Supplemental Fig. 7F). Together, these data suggest that IDO is involved in MAPC cell induction of Tregs, however it is not essential.

CD14 + monocytes are involved in MAPC cell induction of Tregs. MAPC cells and MSC have been
shown to modulate myeloid cell responses by skewing their phenotypic profile towards anti-inflammatory cells or "M2" [54][55][56] . Anti-inflammatory myeloid cells are known to secrete factors such as IL-10, TGFβ, IDO, and retinoic acid that drive Treg differentiation [57][58][59] Considering that monocytes can differentiate into dendritic cells and/or macrophages, and in PBMCs, the percentage of dendritic cells is very low (0.3-0.9% of all leukocytes) while monocytes are more abundant (2-12% of leukocytes), the role of monocytes in MAPC cell induction of Tregs was investigated. To determine whether monocytes are indispensable for MAPC cell induction of Tregs, the percentage of Tregs was assessed in MAPC cell co-cultures with PBMCs in which CD14 + monocytes were www.nature.com/scientificreports/ depleted (Fig. 6). In the absence of CD14 + monocytes, MAPC cells induced Tregs , but, to a lesser extent than unfractionated PBMCs (Fig. 6A). Interestingly, CD14 + monocytes were required for MAPC induction of Ki67 on Tregs (Fig. 6B). Furthermore, while MAPC cells increased the expression of CTLA-4, HLA-DR, and PD-L1 on Tregs induced in the absence of CD14 + monocytes, their expression levels were lower than the unfractionated control (Fig. 6C-E). Supernatant levels of TGFβ and LAP remained unaltered suggesting that MAPC is the primary source of TGFβ and LAP in these cultures and not monocytes ( Supplementary Fig. 8A,B). These data demonstrate that CD14 + monocytes are involved in MAPC cell induction of Tregs.
To further confirm the role of CD14 + monocytes in MAPC cell induction of Tregs, MAPC cell mediated Treg induction was examined in co-cultures of isolated CD4 + T cells and CD14 + monocytes (Fig. 7). Interestingly, MAPC cells significantly increased the percentage of Tregs within isolated CD4 + T cell and CD14 + monocyte cocultures (Fig. 7A). In addition, miTregs had increased proliferation as shown by their Ki67 expression (Fig. 7B). www.nature.com/scientificreports/ Furthermore, their CTLA-4, HLA-DR and PD-L1 expression was also increased (Fig. 7C-E). To assess if cell contact between CD4 + T cells and CD14 + monocytes is required for MAPC cell induction of Tregs, isolated CD4 + T cells were cultured in the well while CD14 + monocytes were co-cultured with MAPC cells in a Transwell membrane (Fig. 7F). Flow cytometric analysis demonstrated that cell contact between CD4 + T cells and CD14 + monocytes is indeed required for MAPC induction of Tregs and Treg proliferation. Collectively, these  Data represent mean ± SD from pooled samples of five independent experiments with 3 PBMC and 3 MAPC cell donors. Statistical analysis was performed using Two-way ANOVA with a Sidak's multiple comparisons test (****p < 0.0001, ***p < 0.001; **p < 0.01, and *p < 0.05). www.nature.com/scientificreports/ data demonstrate that the mechanism by which MAPC cells induce Tregs and promote their proliferation is dependent on CD14 + monocytes. IL-10 is a potent anti-inflammatory cytokine secreted by several immune cells including Tregs, monocytes, macrophages, and dendritic cells. To assess whether MAPC cells affect IL-10 secretion, IL-10 levels in PBMC:MAPC cell co-culture supernatants were measured by ELISA (Fig. 8A). Minimal IL-10 levels were found in supernatants from PBMCs and MAPC cells cultured alone. Interestingly, PBMCs and MAPC cell co-cultures had elevated IL-10 concentration. Analysis of supernatants collected from CD14-depleted: MAPC cell co-cultures demonstrated that IL-10 levels were reduced in the absence of CD14 + monocytes (Fig. 8B). Conversely, IL-10 levels on co-cultures of isolated CD4 + T cells, monocytes, and MAPC cells were comparable to those observed in PBMC:MAPC co-cultures (Fig. 8C), suggesting that the source of IL-10 is either CD4 + T cells or monocytes. To identify the cell responsible for IL-10 production, PBMC:MAPC cells were co-cultured for 7 days. At day 7, GolgiStop was added to the cultures overnight, and IL-10 was measured intracellularly by flow cytometry. IL-10 + cells were found to be CD14 + monocytes and not CD4 + T cells nor MAPC cells (Fig. 8D), confirming monocytes as the primary producer of IL-10 in these cultures. Backgating analysis of IL-10 + cells demonstrated that most of the IL-10 + cells were CD14 + monocytes. Statistical analysis was performed using Two-way ANOVA with a Tukeys's multiple comparisons test (****p < 0.0001 and *p < 0.05). (F) Quantification of Treg percentages and Treg expression of Ki67 after 7-day co-culture of isolated CD4 + T cells with CD14 + monocytes and MAPC cells at 2:1:1 and 2:1:0.5 in direct contact or with MAPC cells and CD14 + monocytes in Transwell. Data represented as mean ± SD from pooled samples of two independent experiments with two different PBMC donors and 2 MAPC cell donors. Statistical analysis was performed using Two-way ANOVA with a Tukeys's multiple comparisons test (****p < 0.0001, ***p < 0.001; **p < 0.01, and *p < 0.05). First, we observed that MAPC cells consistently increased the frequency and total cell numbers of Tregs in a dose-dependent fashion. miTregs expressed higher levels of Foxp3 and CD25 per cell than PBMC Tregs, suggesting that miTregs may be more potent immune regulators than PBMC Tregs, as elevated levels of FoxP3 Tregs have been linked with lower transplant rejection, increased suppressive activity, and higher secretion of IL-10 and TGFβ in a murine model of orthotropic corneal transplantation 32 . Conversely, Tregs expressing low levels depleted PBMCs (gray) cultured alone or with MAPC at 2:1 and 4:1 PBMC:MAPC ratios for 7 days measured by ELISA. Statistical analysis was performed using Two-way ANOVA with a Sidak's multiple comparison test (****p < 0.0001) (C) IL-10 concentration in supernatants from cultures of isolated CD4 + T cells and CD14 + monocytes (2:1 CD4 + cell:CD14 + monocytes) with MAPC cells at 2:1:1 (blue) or 2:1:0.5 (red) measured by ELISA. Data represent mean ± SD from pooled samples of six independent experiments. Statistical analysis was performed using Two-way ANOVA with a Tukeys's multiple comparisons test (****p < 0.0001). (D) Representative histogram depicting intracellular IL-10 expression on MAPC cells (black line), Tregs (blue), and monocytes (red). FMO control shown as dotted line.  14 . MAPC cells induced Tregs to produce low levels of IL-2, cytokine involved in Treg survival, proliferation, and function 68,69 . Given that IL-2 is a strong inducer of Treg proliferation, it is possible that IL-2 secreted by miTregs is involved in the induction of Treg proliferation seen in PBMC:MAPC cell co-cultures.
Phenotypical analysis of cell surface markers demonstrated that miTregs express higher levels of CTLA-4, HLA-DR, and PD-L1 than PBMC Tregs. CTLA-4 is a potent regulator of T cell activation by competing with CD28 for the binding of the B7 costimulatory molecules CD80 and CD86 34 . In humans, Tregs expressing high levels of CTLA-4 have increased suppressive capacity than CTLA-4 low expressing counterparts 37,46 . HLA-DR + Tregs express higher levels of FoxP3 35 . In vitro, activation of HLA-DR − Tregs induces HLA-DR expression 35,43,70 . Parallel comparison of HLA-DR + versus HLA-DR − Tregs demonstrated that HLA-DR + Tregs are more efficient suppressors and HLA-DR expression defines a terminally differentiated Treg 34,35 . MAPC cell mediated induction of HLA-DR was restricted to the Treg population and no other cell types, suggesting that MAPC cells preferentially activate Tregs. Ligation of PD-1 with PD-L1 sends inhibitory signals, thereby downregulating immune activation. Tregs from PD-L1 −/− mice have been shown to have impaired suppressive capacity in vitro and in an in vivo model of nephrotoxic nephritis, suggesting that PD-L1 expression on Tregs is important for Treg suppressive function 71 . Moreover, stimulation of CD4 + T cells with anti-CD3/CD28 in the presence of recombinant PD-L1 increases Tregs by promoting CD4 + T cell differentiation into Tregs and driving Treg expansion 72 . Thus, the increased expression of CTLA-4, HLA-DR, and PD-L1 in miTregs further supports that MAPC cells induce a potent suppressive phenotype on Tregs. Furthermore, miTregs secreted higher levels of TGFβ than Tregs from PBMCs, also suggesting an enhanced suppressive capacity. Using a T cell suppression assay, it was confirmed that miTreg are indeed significantly more suppressive of T cell proliferation when compared to PBMC Tregs. While MAPC cell induction of Tregs has been described in vivo 17,18,24,26 , phenotypic characterization of Tregs induced by MAPC cells in vivo remains to be evaluated.
The CD45 isoforms CD45RA and CD45RO are used to identify resting (CD45RA + FoxP3 + CD25 + ) versus activated (CD45RO + FoxP3 + CD25 + ) Tregs 41,44 . Our studies demonstrated that MAPC cells increase the frequency of a transitional/recently activated Treg expressing both CD45RA and CD45RO. The transitional/recently activated Tregs arose from the CD45RA + Treg population and not from CD45RO + Tregs. Detailed analysis of the phenotype of this transitional population provided evidence that these cells express higher levels of Ki67, correlating with an active proliferative state along with increased CTLA-4 and HLA-DR than the CD45RA + and CD45RO + counterparts, consistent with a recently activated phenotype. MAPC cells also increased the expression of Ki67, CTLA-4, and HLA-DR on both CD45RA + and CD45RO + Tregs; however, consistent with an activated state, CD45RO + Tregs have higher expression of these markers than CD45RA + Tregs.
MAPC cells secrete a variety of paracrine factors that can modulate the immune system including TGFβ and IDO 14,[73][74][75] . The results of our studies demonstrate that TGFβ secreted by MAPC cells is a primary mechanism by which MAPC cells induce Tregs. Blockade of TGFβ signaling using SB 431542, a TGFβ receptor antagonist, significantly abrogated MAPC cell induction of Tregs. In the absence of TGFβ signaling, MAPC cells were less efficient at promoting Treg proliferation and expression of suppressive markers. In addition to TGFβ being a primary driver of MAPC cell induction of Tregs, IDO was also identified as a contributing factor. The inhibition of IDO, another factor secreted by MAPC cells, partially reduced Treg induction supporting IDO involvement in Treg induction by MAPC cells. Both TGFβ and IDO have been extensively identified as key factors secreted by tolerogenic dendritic cells that promote Treg differentiation 45,48,52,53,[57][58][59] . Indeed, TGFβ and IDO have been proposed as potential mechanisms by which MSCs induces Treg differentiation 76,77 . Herein, we demonstrate that these factors also play an important role in MAPC cell mediated Treg induction.
The role of monocytes in MAPC cell induction of Tregs was also explored. Monocytes differentiate into dendritic cells and macrophages and secrete a variety of factors known to promote Tregs. Depletion of CD14 + monocytes demonstrated that these cells are involved in MAPC cell induction of Treg proliferation and expression of CTLA-4, PD-L1, and HLA-DR. In fact, in the absence of CD14 + monocytes, MAPC cells were not able to induce Treg proliferation. These observations were confirmed by co-culture of isolated CD4 + T cells, CD14 + monocytes, and MAPC cells. In this setting, MAPC cells supported Treg induction by increasing their proliferation and expression of a suppressive phenotype. Transwell experiments revealed that direct contact between CD4 + T cells and CD14 + monocytes is necessary for MAPC cell induction of Tregs and Treg proliferation. It is possible that CD14 + monocytes are providing T cell receptor (TCR) stimulation and co-stimulatory signals to Tregs, driving their proliferation. Additional experiments are required to discern the contribution of monocytes in MAPC cell induction of Tregs. The involvement of monocytes in the induction of Tregs by MSCs has been previously described 67,78 . Researchers demonstrated that MSCs skew monocytes towards an anti-inflammatory phenotype, thereby facilitating Treg differentiation 78 . MAPC cells have also been demonstrated to induce a comparable antiinflammatory phenotype on myeloid cells in vitro 55  www.nature.com/scientificreports/ with PBMC. IL-10 has been shown to enhance TGFβ-induced Treg differentiation and suppressive capacity via STAT3 and Foxo1, suggesting that monocyte derived IL-10 might be involved in MAPC induction of Tregs 81 . MAPC cells have potent immunoregulatory properties under a variety of conditions including homeostatic proliferation, graft vs host disease, spinal cord injury, traumatic brain injury and ischemic stroke [14][15][16][17][18][19]61 . Herein, we provide evidence demonstrating that MAPC cells induce Tregs, cells known to play key roles in modulating immune responses in vitro and in vivo, via various mechanisms involving the secretion of TGFβ, IDO, and CD14 + monocytes. Furthermore, characterization of miTregs reveals a more potent immunomodulatory phenotype and highlights a mechanistic pathway where MAPC cells may modulate immune responses under different clinical conditions. Further investigation will be performed in vivo to examine the mechanisms identified in this study involved in MAPC induction of Tregs and perform a functional and phenotypical assessment of miTregs.

Methods
Cell culture. MAPC cells were generated from donor's bone marrow aspirate as previously described 14,82 .
Informed consent was obtained in accordance with the guidelines of a commercial Institutional Review Board for all healthy donors of MAPC cells. Prior to use, MAPC cell phenotype was assessed by flow cytometry. As previously described, all MAPC cells used were over 90% positive for CD49c and CD90, whereas < 5% of the cells expressed HLA-DR and CD45 14 , thereby confirming that the MAPC cells used were in fact a homogenous population. Three different MAPC cell donors were used in this study. The population doublings for all MAPC cells used ranged from 20 to 35.
Human subjects' research approval was obtained from Western Institutional Review Board, Inc. (Puyallup, WA) and written informed consent was obtained from all healthy volunteers involved in this study. All experiments and methods were conducted in accordance with Western Institutional Review Board relevant guidelines and regulations. Peripheral blood mononuclear cells (PBMCs) were isolated from fresh blood of healthy volunteers (ten different donors) using Ficoll-Paque (GE Healthcare Life Sciences, Pittsburgh, PA) density gradient centrifugation as indicated by the manufacturer. Alternatively, isolated frozen PBMCs from healthy donors (two different donors) were purchased from Precision for Medicine (Frederick, MD) and PPA Research (Johnson City, TN). PBMCs (1 × 10 6 ) were co-cultured with MAPC cells at different ratios in RPMI 1640 media supplemented with 10% heat inactivated FBS and 1% penicillin/streptomycin in a 24-well plate. At day 7 post co-culture, supernatants were collected and the percentage of Tregs was determined by flow cytometry.
To identify Tregs, cells were first gated based on lymphocyte size and granularity using forward and side scatter (FSC and SSC), followed by gating on single cells using FSC area versus FSC height. Live cells were selected based on the absence of Fixable Viability stain 780. CD4 + T cells were identified by their expression of CD3 and CD4. Tregs were identified as CD3 + CD4 + FoxP3 + CD25 + (Supplemental Fig. 1). Since CD4 − T cells are known to have low expression of Foxp3 83 , CD4 − T cells were used as a negative control for Foxp3 (Supplemental Fig. 1E).
All samples were analyzed using a BD FACSCelesta flow cytometer equipped with a blue, violet, and red laser configuration. Data collected from these experiments were analyzed using FlowJo software (Ashland, Oregon). Spectral spillover and fluorescent compensation were generated using BDCompBeads sets containing polystyrene beads coupled with antibody specific to mouse, rat, or hamster Ig, κ light chain and negative control beads (BD Biosciences). Positive expression was based on fluorescent minus one (FMO) controls.
T cell suppression assay. To assess Treg suppressive function, isolated PBMCs were cultured alone or with MAPC cells at a 2:1 PBMC:MAPC cell ratio for 7 days as described above. At day 7, Tregs were isolated using EasySep Human CD4 + CD127LowCD25 + Regulatory T cell isolation kit (STEMCELL Technologies, Cambridge, MA) following manufacturer's instructions. Over 85% of the cells isolated were Tregs based on their phenotypic profile (CD3 + CD4 + CD25 + , FoxP3 + , and CD127 low ) assessed by flow cytometry. Isolated Tregs were co-cultured with CellTrace Violet stain (ThermoFisher) labeled autologous PBMCs stimulated with anti-CD3/ CD28 Dynabeads (1 × 10 5 beads/ml, ThermoFisher) in RPMI 1640 media supplemented with 10% heat inactivated FBS and 1% penicillin/streptavidin. At day 5 post stimulation, cells were stained with BD Horizon Fixable Viability stain 780, anti-CD3 PerCP Cy5.5, and anti-CD25 PE (BD Biosciences). Samples were analyzed by flow cytometry using BD FACSCelesta. Tregs were excluded based on their CD25 hi expression and lack of CellTrace Violet stain. T cell proliferation was assessed by measuring CellTrace Violet stain dilution within the CD3 + T cell population. Percent suppression was calculated using the following equation: ELISA. The levels of TGFβ, LAP, IL-2, and IL-10 in culture supernatants were measured using human Quantikine ELISA kits from R&D Systems (Minneapolis, MN) following manufacturer's instructions.

TGFβ and IDO inhibition experiments. To inhibit
Statistical analysis. Data represent the mean ± SD as indicated in figure legends. P-values were determined by One-way ANOVA with a Tukey's multiple comparison test, Student's t Test, or Two-way ANOVA with Sidak's multiple comparison test using GraphPad Prism 8.4.2 (San Diego, CA) as indicated in figure legends.
Ethics approval. The use of healthy donor approval was obtained from Western Institutional Review Board, Inc. (Puyallup, WA) and informed consent was obtained from healthy volunteers as appropriate.

Data availability
The data supporting the findings of this study are available from the corresponding author upon reasonable request.