Adjuvant conditioning induces an immunosuppressive milieu that delays the alloislet rejection through the expansion of myeloid derived suppressor cells

Abstract

Advances in immunosuppression have been relatively stagnant over the last 2 decades and transplant recipients continue to experience long term morbidity associated with immunosuppression regimens. Strategies to reduce or eliminate the dosage of immunosuppression medications are needed. We discovered a novel administration strategy utilizing the classic adjuvant alum to condition murine islet transplant recipients, known as adjuvant conditioning or AC, to expand both polymorphonuclear and monocytic myeloid derived suppressive cells (MDSCs) in vivo. These AC MDSCs potently suppress T cell proliferation when cultured together in vitro. AC MDSCs also facilitate naïve CD4⁺ T cells to differentiate into regulatory T cells. In addition, we were able to demonstrate a significant delay in alloislet rejection compared to saline-treated control following adjuvant treatment in a MDSC dependent manner. Furthermore, AC MDSCs produce significantly more IL-10 compared to saline-treated controls, which we demonstrated to be critical for the increased T cell suppressor function of AC MDSCs, as well as the observed protective effect of AC against alloislet rejection. Our data suggest that adjuvant related therapeutics designed to expand MDSCs could be a useful strategy to prevent transplant rejection and curb the use of toxic immunosuppressive regimens currently employed in transplant patients.

1. Introduction

While transplant allograft and patient survival have improved, potent immunosuppressive medication regimens continue to cause a significant amount of morbidity in transplant recipients. Based on the most recent Scientific Registry of Transplant Report, allograft survival following solid organ transplant is 93%, 80%, and 50% for deceased donor kidney transplants and 91%, 78%, and 57% for deceased donor liver transplants at 1-, 5- and 10-years post-transplant, respectively^1,2. While this ultimately translates into an increased lifespan of transplant recipients, it also results in longer exposure to high levels of immunosuppressive regimens including steroids and calcineurin inhibitors like tacrolimus during both the induction phase (perioperatively) and the maintenance phase (months to years following)³. Although these regimens are critical for allograft survival, they also cause significant morbidity including increased susceptibility to infections, chronic kidney injury, and cancer ⁴. Therefore, strategies to reduce or eliminate these regimens are needed.

Previous efforts to promote allograft tolerance have focused on controlling adaptive immune responses (i.e., through T regulatory (Treg) cell induction); however, there is still much about the role of the innate immune system in alloimmunity that is unknown ^5,6. Myeloid derived suppressor cells (MDSCs) are a heterogeneous population of innate immune effector cells that arise in multiple physiologic/disease states. MDSCs are powerful suppressors of the adaptive immune response but, importantly, preserve immunity to infection and sepsis ^5,7. MDSCs are a heterogeneous innate immune population that is composed of both immature granulocytes and polymorphonuclear (PMN) cells (CD11b⁺Ly6G⁺) (PMN-MDSCs) and monocytes (CD11b⁺Ly6G⁻Ly6C⁺) (M-MDSCs) ⁵. In addition to the many mechanisms of MDSC-dependent immunosuppression that have been described, MDSCs have also been shown to coordinate T, innate lymphoid cells, and B regulatory cell expansion^8-10.

Inflammation has been shown to be an important physiologic driver of MDSC expansion ^5,11. In addition, inflammation has been shown to be a critical component of adjuvant and vaccine administration through the production of cytokines and chemokines that recruit and/or activate innate immune cells leading to adaptive immune cell activation and memory formation. Despite this, Schmoekel et al. has demonstrated that mice administered an adjuvant multiple times are characterized by an immunoinhibitory response ¹². Although Treg cells may play a role in this process, it remains unknown whether MDSCs could also play a role.

Through a novel administration strategy of the classic adjuvant aluminum hydroxide (alum), hereafter termed adjuvant conditioning (AC), we are able to expand both immunosuppressive PMN and monocytic MDSCs. We also demonstrate that AC MDSCs expand T regulatory T cells (Tregs) in vitro compared to Gr-1⁺ cells from saline-treated animals. In addition, we observe a significant delay in alloislet rejection compared to saline-treated controls. We also demonstrate that the observed protective effect of AC is dependent on MDSCs. Furthermore, this protective effect can be adoptively transferred as AC MDSCs administered to murine islet recipients can also delay the rejection of alloislets. We also show that AC-elicited MDSCs produce higher amounts of IL-10 compared of naïve counterparts and that blockade of IL-10 in vitro compromised the T cell suppressive ability of MDSCs and restored T cell proliferation. In keeping with our in vitro results, we find that in vivo administration of an anti-IL10 monoclonal antibody (mAb) abolished the protective effects of AC. Finally, we demonstrate that depletion of Tregs via anti-CD25 administration has no effect on AC mediated protection of alloislets, further supporting the role of MDSCs in our model. Altogether, these data demonstrate a novel strategy to reduce alloimmunity through adjuvant conditioning via the induction of potent immunosuppressive MDSCs.

2. Materials and Methods

2.1. Mice

8-10-week-old male C57BL/6J (B6, H-2^b), Balb/C (B/c, H-2^d) and OTI mice were purchased from Jackson Laboratory (Bar Harbor, ME). All mice were kept under specific pathogen free facility at Boston Children’s hospital. All mouse experimental protocols were approved by the Institutional Animal Care and Use Committees of Boston Children’s Hospital.

2.2. Drug administration

For adjuvant/alum administration, mice received either saline or alum 200 mL intraperitoneal (IP) (Sigma-Aldrich) depending on the experimental schematic described in each figure. For IL-10 blockade, anti-IL-10 antibody (JES5-2A5, Bio X cell) was injected IP at a dose of 200 μg/mouse at time points described below. For MDSC depletion, 250μg anti-Ly6G (1A8, Bio X Cell) or anti-Gr1 (RB6-8C5, Bio X Cell) were injected IP¹³. For Treg depletion, 200μg/mouse of anti-CD25 (PC-61.5.3, Bio X cell) was injected IP as indicated as previously described¹⁴.

2.3. Islet transplant

To perform the islet transplant model, alloislet recipient mice (C57BL/6J) were injected with streptozotocin (STZ) 240mg/kg IP to induce diabetes (Sigma-Aldrich). The induction of diabetes was then confirmed in each mouse if blood glucose was >250 mg/dL for 3 consecutive days. Islets from donor mice (Balb/C) were isolated by the standard technique via pancreas collagenase digestion and density gradient centrifugation as described in prior studies ¹⁵. 250 to 300 islets were then transplanted under the renal capsule of recipient mice. Euglycemia or blood glucose level < 200 mg/dL was then monitored for 3 consecutive days post-transplant to ensure technical success of alloislet transplantation. Rejection was defined when blood glucose >250 mg/dL for 2 consecutive days.

2.4. Cell enrichment and adoptive transfer

Gr-1⁺ cells were isolated via negative selection from mouse spleens using the EasySep mouse MDSC (CD11b⁺Gr1⁺) isolation kit (StemCell) according to manufacturer’s protocol. Purity of isolated MDSCs were verified via flow cytometry to be greater than 92%. For adoptive transfer experiments, 5x10⁶ of purified MDSCs were injected intravenously at the day 0, 7, and 14 days post islet transplant as indicated below.

2.5. Flow cytometry

Single cell suspensions were prepared from peripheral blood and splenocytes after RBC lysis using ACK lysis buffer (ThermoFisher Scientific). The following antibodies were used for surface staining: CD11b (Clone: M1/70, Invitrogen), Ly6C (Clone: HK1.4, BioLegend), Ly6G (Clone: 1A8, BioLegend), CD3 (Clone: 17A2, Invitrogen), CD4 (Clone: GK1.5, Invitrogen), CD8 (Clone: 53-6.7, BioLegend), CD44 (Clone: IM7, BioLegend), CD62L (Clone: MEL14, BioLegend), PD1 (Clone: RMP1-30, BD Biosciences), PDL1 (Clone:10F.9G2, BioLegend), CD115 (AFS98, BioLegend), IL-4R (Clone: I015F8, BioLegend). Samples were fixed and permeabilized with Fix/Perm buffer according to manufacturer’s instruction (eBioscience, San Diego, CA) before intracellular protein staining. The following antibodies were used for intracellular staining. IL-10 (Clone: JES5-16E3, BioLegend), FoxP3 (Clone: MF-14, BioLegend), IFN-γ (Clone: AN-18, BioLegend). PMA/Ionomycin (Sigma-Aldrich) and GolgiPlug (BD Biosciences) were cultured with T cells for 5 hours for IFN-γ staining. BioLegend LEGENDplex multiplex assay was used for serum cytokine assessment. Serum samples from mice were processed with Mouse Inflammation Panel kit according to manufacturer’s instructions. Samples were collected on LSR Fortessa HTS (BD) with BD FACSDiva v8.0.2 software. FlowJo v10 was used for flow data analysis.

2.6. In vitro suppression assay

Cells were cultured in RPMI 1640 medium (ThermoFisher Scientific) containing 10% FBS (Corning), 200 μg/mL penicillin (ThermoFisher Scientific), 200 U/mL streptomycin (ThermoFisher Scientific) and 0.05mM 2-mecaptoethanol (Sigma). For MDSC suppression assay, naïve CD4⁺ T cells were isolated from mouse spleens using EasySep mouse naïve CD4⁺ T cell isolation kit (StemCell) and labeled with CellTrace Violet (CTV, ThermoFisher Scientific). Gr-1⁺ cells were isolated from alum or saline-treated animals by EasySep mouse MDSC (CD11b⁺Grl⁺) isolation kit (StemCell) and cultured with naïve CD4⁺ T cells in the presence of IL-2 (1ng/ml) and CD3/CD28 activation beads (ThermoFisher Scientific) for 72 hours. To measure MDSC induced suppression of antigen-specific T cell proliferation, OTI CD8⁺ T cells isolated from OTI mice spleens labeled with CTV were cultured with M-MDSCs and PMN-MDSCs purified by MACS separation column (Miltenyi) and OVA_257-264 peptide (2μg/mL, Invivogen) with naïve splenocytes for 48 or 72 hours as indicated. 10ng/ml anti-IL10 antibody (JES5-2A5) (Bio X cell) was added to the culture to neutralize IL-10. CTV dilution was measured by flow cytometry to identify cell proliferation.

2.7. Treg induction assay

Naïve CD4⁺ T cells isolated from C57Bl/6 spleens via EasySep mouse naïve CD4⁺ T cell isolation kit (StemCell). M-MDSCs or PMN-MDSCs isolated from naïve or alum conditioned mouse spleens using myeloid-derived suppressor cell positive isolation kit (Miltenyi). Briefly, splenocytes were labeled with anti Ly6G-biotin antibody following FC receptor blockage. After incubated with streptavidin conjugated MicroBeads, Ly6G⁺ cells were selected through MACS Column system. The unlabeled flow-through fraction (Ly6G⁻ cells) underwent Ly6C-biotin antibody labelling followed by magnetic labelling and column isolation. MDSC purity was examined by flow cytometry. Then either M-MDSCs or PMN-MDSCs were cultured with naïve CD4⁺ T cells in complete medium with IL-2 (1ng/ml) for 72 hours as described previously¹⁶. Cells suspensions were stained for CD4⁺ FoxP3⁺ and analyzed by flow cytometry after co-culture.

2.8. Data Analysis and Statistics

Statistical analysis was performed using GraphPad Prism version 9. Two-tailed unpaired Student’s t test was used to calculate differences between experimental animals. One-way analysis of variance (ANOVA) was used for multiple comparisons. Graft survival significance was assessed by Kaplan-Meier/Mantel-Cox log-rank test. P value < 0.05 was considered to be a statistically significant different.

3. Results

3.1. Adjuvant administration induces the expansion of MDSCs.

Schmoekel et al demonstrated that non-specific adjuvant administration can suppress the production of ovalbumin specific IgG through Treg cell expansion¹². However, the involvement of MDSCs was not explored. Since inflammation induces the expansion of MDSCs, we explored whether administration of the classic adjuvant alum, could also expand monocytes and granulocytes with the features of MDSCs ^5,11,17. Though there was a limited expansion of splenic granulocytes, CD11b⁺Ly6G⁺ cells, following one dose of adjuvant (Supplemental Figure 1A), we found that 3 doses of alum, delivered every other day, had the most profound effect on both granulocyte and monocyte expansion in the peripheral blood and spleen compared to saline controls (Supplemental Figure 1A and Figure 1A and B). To distinguish multiply adjuvant dosed mice from mice in which a single dose of adjuvant was given, we termed the administration of adjuvant multiple times over the course of 5 days “adjuvant conditioning” or AC. In addition to the expansion of myeloid cells, AC-treated mice also developed mild splenomegaly that is seen in other murine models of inflammation, sepsis, or cancer secondary to the expansion of myeloid cells (Supplemental Figure 2)^18,19.

FIGURE 1. Adjuvant conditioning (AC) expands MDSCs in vivo with immunoregulatory features.

(A) Experimental Schema: C57BL/6J mice were injected with 200 μL alum (AC) or saline IP every other day for 3 doses. (B) Adjuvant conditioning (AC) induces monocytic and granulocytic myeloid cell expansion in both peripheral blood and spleens (n=4-7). (C) Myeloid cells from AC treated mice suppress CD4⁺ T cell proliferation in vitro. CTV labelled naïve CD4⁺ T cells were stimulated with anti CD3/CD28 activation beads and IL-2 (1ng/ml) and cultured with or without splenic Gr-1⁺ MDSCs isolated from alum or saline-treated mice at a ratio of 1:2 (Gr1⁺: CD4). Representative flow cytometry figures of CD4⁺ T cells proliferation (left) and comparison of proliferated CD4⁺ T cells (right). (D) Myeloid cells from AC treated mice suppress antigen-specific CD8⁺ T cells proliferation in vitro. CD11b⁺ Ly6C⁺ M-MDSCs or CD11b⁺ Ly6G⁺ PMN-MDSCs isolated from alum or saline-treated mice spleens were cultured with. Experiments in (C) and (D) were performed as technical triplicates and repeated at least twice. Data are expressed as mean ± SEM; student’s t test was used for analysis; p-values are listed above experimental groups being compared.

To further characterize these myeloid cells, we performed cell surface staining for markers typically associated with myeloid derived suppressor cells (MDSCs), namely PD-L1, CD115 (M-CSFR), and IL-4R²⁰. Compared to saline treated controls, AC increased the expression of PD-L1 and IL-4R but had no effect on CD115 expression (Supplemental Figure 3A and 3B). Since these findings mirror MDSC expansion in other pathologic/physiologic states, we hypothesized these expanded myeloid cells could suppress adaptive immune responses, a critical component of MDSC function⁷. To test this hypothesis, we cocultured adjuvant expanded myeloid cells with either T cells stimulated with anti CD3/CD28 beads or with ova specific CD8⁺ T cells (OT-I), and OVA peptide. AC expanded myeloid cells were able to suppress T cell responses under both experimental conditions (Figure 1D, E). These data support the observation that AC expanded myeloid cells were likely MDSCs (AC MDSCs).

3.2. Adjuvant conditioning suppresses the rejection of alloislets.

Following this observation, we wanted to determine whether AC could suppress allograft rejection. To this end, we utilized a classic transplant model in which Balb/C islets were transplanted into streptozotocin-induced diabetic C57Bl/6J mice that were adjuvant conditioned or saline-treated (Figure 2A). Due to the fact that the onset of diabetes (blood glucose greater than 250mg/dL on 2 consecutive measurements) correlated, with significant moribund changes in the mouse habitus, the onset of diabetes was considered as a survival endpoint. As demonstrated in Figure 2B, adjuvant conditioned animals experienced a significant delay in the development of diabetes and the rejection of fully mismatched alloislets compared to saline-treated controls. This is in contrast to the observed survival of mice following alloislet transplantation in mice administered one dose of adjuvant (Supplemental Figure 1B). As a results, survival of AC mice was significantly prolonged compared to controls.

FIGURE 2. Adjuvant conditioning (AC) prolongs islet allograft survival.

(A) Experimental Schema: C57BL/6J mice received either AC or saline IP every other day for 3 doses before Balb/C alloislet transplant as indicated. STZ was given at −6 days of islet transplant to induce diabetes. Mice were monitored for graft rejection via plasma glucose measurement. (B) Survival curve of STZ treated mice that received either AC or saline administration followed by alloislet transplant. Statistical analysis was performed using the log-rank test.

3.3. Adjuvant administration of MDSCs induces Treg expansion in the spleen and draining lymph nodes.

Since Tregs play important roles in suppressing alloimmune responses, we hypothesized that adjuvant conditioning could expand Tregs. To test this hypothesis, AC or saline-treated streptozotocin-induced diabetic C57Bl/6J mice underwent islet transplants from Balb/C mice (Figure 3A). Ten days after transplant, spleens and the left renal draining lymph nodes (LN) were assayed for the presence of Tregs (FoxP3⁺). As demonstrated in Figure 3B, there was a significant expansion of Tregs in AC-treated mice vs saline-treated mice following islet transplant. Splenic and draining LN-isolated Tregs from AC-treated animals also had significantly higher expression of the suppressive immune checkpoint molecule program cell death protein 1 (PD-1) than saline-treated controls following transplant (Figure 3C). We also examined the phenotype of non-FoxP3+ T cells in the draining lymph node and spleen. We observed decreased IFN-γ producing T cells in the spleens and in the peripheral blood upon AC treatment (Supplemental Figure 4A and 4B). In addition, ten days post islet transplant, there were more naïve T cells (CD44-CD62L+) and fewer activated/memory phenotype T cells with AC treatment compared to saline treated controls (Supplemental Figure 4C).

FIGURE 3. AC MDSCs induce Treg differentiation.

(A) Experimental Schema: C57BL/6J mice received either AC or saline IP prior to undergoing Balb/C alloislet transplant. Draining lymph nodes (dLN) from left kidneys (with alloislet grafts) and spleens (Sp) were collected to assess for Tregs via flow cytometry. The ratio of FoxP3⁺ cells of CD4⁺ T cells (B) and PD1⁺ Tregs (C) were analyzed. (D) CD11b⁺ Ly6G⁺ PMN-MDSCs or (E) CD11b⁺ Ly6C⁺ M-MDSCs or isolated from saline or AC treated mice spleens were cultured with naïve CD4⁺ T cells for 72h. FoxP3⁺ CD4⁺ T cells were assessed with flow cytometry following co-culture. Experiments were repeated twice in triplicates. Data are expressed as mean ± SEM; student’s t test was used for analysis; p-values are listed above experimental groups being compared.

Previous studies have demonstrated that MDSCs can facilitate the expansion of Tregs ^9,21. To explore whether AC MDSCs can differentiate Tregs from naïve CD4⁺ T cells in vitro, AC MDSCs or myeloid cells from saline treated controls were cultured in vitro with naïve CD4⁺ T cells. We found a significantly greater expansion of FoxP3⁺ Tregs when naïve T cells were co-cultured with AC MDSCs, compared to saline controls (Figure 3D). These data suggest that adjuvant conditioned myeloid cells can facilitate the expansion of immunosuppressive FoxP3⁺ T cells.

3.4. Deletion of Ly6G⁺ or Gr1⁺ cells abrogates the protective effect of adjuvant elicited MDSCs on allograft rejection.

Our data demonstrate that AC MDSCs efficiently suppress T cell activation and favor Treg expansion in vitro. Next, we investigated whether the activity of AC-induced myeloid cells was required to delay the rejection of fully mismatched islets in our transplant model. To examine this, alum-conditioned animals were administered either anti-Gr-1 or anti-Ly6G antibodies as shown in the schematic (Figure 4A). Depletion of Gr-1⁺ or Ly6G⁺ cells via Ly6G or Gr-1 was confirmed by flow cytometry (Supplemental Figure 5). Following depletion of either Gr-1⁺ or Ly6G⁺ cells, we found that the protective effect of adjuvant conditioning was abrogated (Figure 4B and 4C). These data suggest that AC MDSCs, and specifically the CD11b⁺Ly6G⁺ population, are critical to delay alloislets rejection.

FIGURE 4. Depletion of Ly6G population in AC mice abrogates prolongation of alloislet survival.

(A) Experimental Schema: STZ-conditioned diabetic C57BL/6J mice received Balb/C islets following AC or saline administration. In a separate cohort of mice, anti-Ly6G (250μg/mouse) or anti-Gr1 (250μg/mouse) depletion antibodies were injected IP at day −4 and −2 of islet transplant for in vivo MDSC depletion. Isotype antibodies for anti-GR1 or anti-Ly6G were used for controls. Mice were monitored for graft rejection via plasma glucose measurement. Survival curve depicts islet graft survival of mice that were treated with either (B) anti-GR1/isotype control or (C) anti-Ly6G/isotype control administration. Statistical analysis was performed using the log-rank test (*p= 0.019 **p = 0.0495).

3.5. Adoptive transfer of adjuvant elicited MDSCs confers a protective effect

To further investigate the potential role for AC-expanded MDSCs in the protection against alloislets rejection, streptozotocin induced diabetic mice were transplanted with Balb/C islets and then subsequently adoptively transferred with 5 x 10⁶ MDSCs at weekly intervals following transplant (Figure 5A). Animals that were administered AC MDSCs had a significant delay in the rejection of fully mismatched islets compared to those administered Gr-1⁺ cells from naïve mice (P=0.0251) (Figure 5B).

FIGURE 5. AC MDSCs prolong alloislet graft survival upon adoptive transfer.

(A) Experimental Schema: C57BL/6J mice underwent Balb/C alloislet transplant following STZ administration as indicated. Next, either saline versus 5x10⁶ splenic Gr-1⁺ cells isolated from AC or saline-treated mice were injected intravenously at day 0, 7, and 14 post islet transplant as indicated. Mice were monitored for graft rejection via plasma glucose measurement. (B) Survival curve of alloislet transplanted mice treated with either saline vs adoptively transferred with CD11b⁺ GR-1⁺ cells from mice treated with either AC or Saline. Statistical analysis was performed using the log-rank test.

3.6. Blockade of IL-10 partially reverses the immunosuppressive effect of adjuvant elicited MDSCs.

As our data have shown, AC MDSCs are able to expand Tregs in vitro and Tregs are increased in the draining lymph nodes of AC alloislet transplant recipients. As MDSCs have been shown to engage in cross talk with Tregs and Treg expansion and function has been demonstrated to be augmented by IL-10, we sought to investigate whether IL-10 expression was important in our observations^21,22. First, to dissect the immunosuppressive mechanism by which AC MDSCs suppress alloimmunity, we initially investigated in vitro two well-known mechanisms that suppress T cell responses: program cell death protein ligand 1 (PD-L1) and IL-10. As shown in Figure 6A, AC MDSCs produced significantly more IL-10 compared to naïve myeloid CD11b⁺Ly6C⁺ or Ly6G⁺ cells. Consistent with this, IL-10 concentration from plasma was increased following AC treatment (Figure 6B). As IL-10 could not be detected in CD45⁺ CD11b⁻ or CD11b⁺Ly6C⁻Ly6G⁻ populations (Supplemental Figure 6A and 6B), we reasoned that only AC MDSCs were responsible for this IL-10 level alteration. Though blockade of PD-L1 had no effect (data not shown), in vitro blockade of IL-10 partially reversed the suppressive capacity of AC MDSCs (Figure 6C). To further investigate this effect in vivo, we investigated the result of IL-10 blockade on allograft rejection in our model. As shown in Figure 6D and 6E, we demonstrate complete loss observed protective effect of adjuvant conditioning in animals that were treated with anti-IL10 versus those treated with isotype control. These data suggest that an IL-10 MDSC dependent alloimmune suppression is critical for the observed delay in rejection in adjuvant conditioned mice.

FIGURE 6. AC MDSCs produce IL-10 to prolong alloislet graft survival.

(A) PMN-MDSCs or M-MDSCs from AC or saline mice were assessed for IL-10 production via flow cytometry (left). Mean fluorescence intensity (MFI) fold change relative to isotype flow antibody control were compared (right). Data are expressed as mean ± SEM; student’s t test was used for analysis; p-values are listed above experimental groups being compared. (B) AC leads to increased IL-10 concentration in plasma. Student’s t test was used for analysis. Serum samples obtained from LPS conditioned mice (6 hours after LPS 10mg/kg IP) were used as positive controls (* p< 0.01; ** p<0.001). Experiments were performed in triplicate and were compared using student’s t test; p-values are listed above experimental groups being compared. (C) Blocking of IL-10 partially reverses AC MDSCs mediated T cell suppression. CD11b⁺ Ly6C⁺ M-MDSCs or CD11b⁺ Ly6G⁺ PMN-MDSCs splenocytes were isolated from AC treated mice were cultured with OTI CD8⁺ T cells in the presence of OVA peptide and naïve splenocytes. Anti-IL10 (10 ng/ml) was used for IL-10 neutralization. OTI T cell proliferation was assessed by flow cytometry after cells were cultured for 48 hours (left) or 72 hours (right). Experiments were performed in triplicate and were compared using student’s t test; p-values are listed above experimental groups being compared. (D) Experimental Schema: C57BL/6J mice received Balb/C islets following AC and received 5 doses of anti-IL10 mAb (200 μg/mouse) every other day. Mice were monitored for graft rejection via plasma glucose measurement. (E) Survival curve of mice receiving anti-IL10 or isotype mAb. Statistical analysis was performed using the log-rank test.

3.7. T regulatory cells are not critical for the protection observed following adjuvant conditioning.

Tregs are thought to be key mediators of peripheral tolerance and to do so through cell-cell contacts as well to utilize paracrine means, such as the production of IL-10. Therefore in our model, we sought to understand the role of Tregs in the protection of alloislets following AC. As demonstrated above, depletion of MDSCs and in particular the granulocyte compartment had the most profound effect on the observed protection of MDSCs (Figure 4). As MDSCs have been shown to be important in the expansion of Tregs, we wanted to examine if there was any effect on the presence of splenic Tregs following MDSC depletion with anti-Gr-1. Surprisingly, there was no effect on Treg expansion (Figure 7A and 7B). To further elucidate this, we examined what the role, if any, Tregs were playing in the observed protective effect of AC on alloislet rejection. Using a fairly well-established method to deplete Tregs via anti-CD25 administration, we did not observe any differences between those animals anti-CD25 vs isotype controls (Figure 7C and 7D). While surprising, these data again highlight the importance of MDSC dependent IL-10 production in the observed protected effect on alloislet rejection seen following AC.

FIGURE 7. T regulatory cells are not critical for the protection observed following adjuvant conditioning.

(A) Experimental schema (B) Percentage of CD4⁺FoxP3⁺ in draining lymph node 10 days post alloislet transplant (n=3 per group) (not significant = NS). Student’s t test was used for analysis; p-values are listed above experimental groups being compared. (C) Experimental schematic: Anti-CD25 mAb (200 μg IP at day −2 and 2 of alloislet transplant) was administrated to AC treated mice for in vivo Treg depletion. (D) Survival curve of AC recipients received either anti-CD25 or isotype control antibodies. Mice were monitored for graft rejection via plasma glucose measurement. Statistical analysis was performed using the log-rank test.

4. Discussion

Myeloid-derived suppressor cells have been shown to play critical roles in adaptive immunosuppression in multiple disease models including cancer, autoimmunity, and sepsis ^5,7,23. The mechanism by which MDSCs suppress adaptive immunity has been ascribed to a number of pathways including metabolic (arginase), paracrine (production of reactive oxygen species and cytokines) and direct cell-cell interaction with immunoinhibitory cell surface markers (PD-L1) ^7,24. MDSCs have also been shown to coordinate the responses of other immunoregulatory cells such as Tregs, Bregs, and innate lymphoid cells^8,22. While their deleterious role in malignancy and other disease pathology is well described, several groups have demonstrated a potential therapeutic role for MDSCs in transplantation, but their application in controlling alloimmunity remains largely underutilized and undeveloped ^6,8,23,25. In this report, we present a novel strategy by using the adjuvant alum to condition murine transplant recipients prior to allograft transplant termed, adjuvant conditioning (AC).

Despite the fact that both monocytic and PMN MDSC populations are expanded following AC, PMN-MDSCs have increased immunosuppressive activity in vitro through the suppression of antigen-specific T cell proliferation. In addition, when the expansion of Ly6G expressing cells (granulocytes and PMN-MDSCs) following AC is blocked in vivo, the protective effect in AC mice is lost. These data suggest that MDSCs, and specifically the PMN-MDSC population, are a critical component of the observed protective effect of AC.

While MDSCs have been demonstrated to express immunoinhibitory cell surface markers, including PD-L1 and paired immunoglobulin-like receptor B (PIR-B), as well as secrete many paracrine factors that suppress adaptive immunity, we identified IL-10 as being the most important in our model system ²⁶. We demonstrated that AC expanded MDSCs produce significantly more IL-10 than saline-treated controls. In addition, we also observed that blockade of IL-10 in AC-treated animals abrogates the protective effect of AC on alloislet transplantation. Also, we have also demonstrated that blockade of Tregs has no effect on the observed protective effect of AC on alloislet rejection. Certainly, we cannot exclude the possibility that other MDSC-dependent factors may play a role in the observed immunosuppressive effect. Despite this, we have demonstrated that IL-10 and possibly MDSC dependent IL-10 production is critical for the alloregulatory response in observed in our system. Furthermore, although not as critical for the observed effect of AC, MDSC-dependent Il-10 is also exerting its effect via well-established role through the augmentation of Treg function, expansion of Tregs in the draining LN, and/or direct inhibition of alloreactive T cells ^27,28. Whether MDSCs are coordinating Treg responses in our system is unknown but also not important for the observed effect of AC on alloislet rejection.

In both cancer and transplant models, MDSCs have been shown to engage in crosstalk with other immunoregulatory cells such as Tregs through either direct interaction via the PDL1 pathway or through soluble factors such as IL-10 ^21,22,29,30. We observed a significant increase in T regulatory cells (Tregs) and MDSCs in the draining lymph nodes of AC-treated animals that were transplanted with alloislets compared to saline-treated controls. In addition, when AC-treated PMN-MDSCs, but not M-MDSCs, were incubated in vitro with naïve CD4⁺ T cells, we demonstrate an increase in the number of FoxP3⁺ Tregs. Depletion of Gr-1 positive cells demonstrated a modest but not significant decrease in draining lymph node Tregs (Figure 7B). These data are supported by findings in other studies suggesting that MDSCs can indeed support or expand Tregs and while Tregs were found to not be critical for the observed effect of AC on alloislet rejection the observation that Treg expansion does occur in the model is not deleterious and may be important to establish longer lived allografts.

One potential limitation of this study is the restricted long-term effect of AC treatment on the survival of islets. Despite attempts to administer booster injections of adjuvant at Day 7 post-transplant to potentially prolong the protective response of AC, the survival of the alloislet grafts was the same (data not shown). Although we did not attempt additional administrations of adjuvants to be helpful in this study, our findings are akin to what was described by Lee et al. in that the role of MDSCs seemed to be critical for the early protection of heterotopic heart allografts²³. As antigen specificity is thought to be important for the creation and maintenance of tolerance, this may also explain why Tregs may not be as important in our model. Certainly, investigations including relevant donor antigens in the conditioning regimen would be very exciting and are the scope of future studies. Although additional strategies are needed to prolong the life of the allograft long term, these observations are important in the generation of therapeutics designed to expand alloregulatory MDSCs and will guide design of future therapeutics to modulate alloimmunity.

In conclusion, we have demonstrated that adjuvant conditioning can be used to significantly prolong the survival of alloislets through the expansion of alloregulatory MDSCs. These MDSCs were important for the early suppression of alloimmunity and for the rejection in our model which was Treg independent. While other therapeutics may be needed to provide long-term survival, the use of MDSCs can be useful to decrease our dependence on potent induction regimens such as steroids and thymoglobulin which can negatively affect not only MDSCs but other alloregulatory populations ^31,32.

Supplementary Material

Supplemental Figure 1

Supplemental Figure 1. Expansion of splenic granulocytes and alloislets survival following one time adjuvant administration. (A) Expansion of CD11b⁺Ly6G⁺ myeloid cells post saline, one-time alum administration, and alum administration three times every other day. (B) Survival of mice post alloislet transplantation and adjuvant administration.

Supplemental Figure 2

Supplemental Figure 2. AC causes splenomegaly (saline: n=6, AC: n=8)

Supplemental Figure 3

Supplemental Figure 3. Expression of PDL1, CD115, and IL-4R in splenic myeloid cells isolated from saline vs AC treated mice. Flow cytometry plots are representatives of CD11b⁺Ly6C⁺ (A) or CD11b⁺Ly6G⁺ (B) cells from animals with either saline or AC treated groups that express PD-L1, CD115, and IL-4R. Bar graphs represent percentage of cells that express PD-L1, CD115, and IL-4R in either saline or AC treated groups. (C) PDL1 expression in AC M-MDSCs or AC PMN-MDSCs. Graph shows the percentage of PDL1 positive cells from each group. Statistical analysis was performed using student’s T-test.

Supplemental Figure 4

Supplemental Figure 4. T cell effector functionality was impaired upon AC in vivo (A) representative plots (A) and percentage (B) of IFN-γ expression in T cells from spleen or peripheral blood of mice treated with saline or AC. (C) Comparison of memory (CD44⁺CD62L⁺), effector (CD44⁺CD62L⁻), naïve (CD44⁻CD62L⁺) T cells from draining lymph nodes in islets graft recipients with saline or AC treatment. Experiments were repeated twice with 3 animals in each group. Student’s T-test was used for statistical analysis.

Supplemental Figure 5

Supplemental Figure 5. Demonstration of myeloid cell depletion following isotype, anti-Gr-1, or anti-Ly6G administration. Depletion or isotype control antibodies were injected IP at day −4 and −2. Splenocytes were analyzed through flow cytometry at day 1 after antibody treatment.

Supplemental Figure 6

Supplemental Figure 6. IL-10 was upregulated exclusively by myeloid cells upon AC treatment. (A) Representative plot of IL-10 expression of splenocytes and peripheral blood from saline and AC treated mice. Plots were gated on CD45⁺ live cells. (B) IL-10 expression of CD11b⁺ Ly6G⁻Ly6C⁻ cells.

Abbreviations:

AC

adjuvant conditioning

Breg

regulatory B cells

IL

interleukin

IP

intraperitoneal

mAb

monoclonal antibody

MDSCs

myeloid derived suppressive cells

M-MDSCs

monocytes myeloid derived suppressive cells

OVA

ovalbumin

PMA

phorbol myristate acetate

PMN-MDSCs

polymorphonuclear myeloid derived suppressive cells

RBC

red blood cell