In vivo expansion of antigen-specific regulatory T cells through staggered Fc.IL-2 mutein dosing and antigen-specific immunotherapy

Abstract

In mice, antigen administration in the absence of adjuvant typically elicits tolerogenic immune responses through the deletion or inactivation of conventional CD4 T cells (Tconv) and the formation or expansion of regulatory CD4 T cells (Treg). While these “antigen-specific immunotherapy” (ASI) approaches are currently under clinical development to treat autoinflammatory conditions, efficacy and safety may be variable and unpredictable due to the diverse activation states of immune cells in subjects with autoimmune and allergic diseases. To reliably induce antigen-specific tolerance in patients, novel methods to control T cell responses during ASI are needed; and strategies that permanently increase Treg frequencies among antigen-specific CD4 T cells may provide long-lasting immunosuppression between treatments. Here, we present an approach to durably increase the frequency of antigen-specific Treg by administering ASI when Treg numbers are transiently increased with individual doses of a half-life-extended Treg-selective IL-2 mutein. Repeated weekly cycles of IL-2 mutein doses (day 0) followed by ASI (day 3) resulted in a 3- to 5-fold enrichment in Treg among antigen-responsive CD4 T cells. Expanded antigen-specific Treg persisted for more than 3 weeks following treatment cessation, as well as through an inflammatory T cell response to an antigen-expressing virus. Combining Treg enrichment with ASI has the potential to durably treat autoimmune disease or allergy by increasing the Treg/Tconv ratio among auto-antigen- or allergen-specific T cells.

INTRODUCTION

Immune homeostasis is mediated in part through a tug-of-war between pro- and anti-inflammatory T cells, in which populations of self- and foreign-antigen-reactive T cells are chronically restrained by immunosuppressive Treg. In autoimmunity, self-reactive Tconv overpower Treg, resulting in detrimental attack of self-tissues; and one approach to restore balance is to boost the quantity and quality of Treg relative to Tconv (1). Based on extensive preclinical work, enhancing Treg in autoimmune patients is currently being tested through two main approaches: 1) stimulation of Treg growth and function via therapeutic administration of recombinant IL-2 or IL-2 variants (muteins) (2, 3), and 2) adoptive transfer of in vitro expanded patient-derived Treg (4, 5). While reports of clinical efficacy in autoimmune diseases are forthcoming for IL-2 muteins and adoptive Treg therapy, robust and consistent reversal of autoimmunity with low-dose IL-2 has not yet been demonstrated (6–8). If these approaches ultimately fail to reach clinical trial endpoints, poor efficacy may result from 1) weak Treg agonism and/or poor Treg-selectivity with low-dose IL-2 or IL-2 muteins, 2) induction of IL-2R non-responsiveness following extended IL-2 or IL-2 mutein treatment, 3) poor persistence of adoptively transferred Treg, and/or 4) low representation of Treg specific for disease-associated antigens within enhanced Treg populations. Regarding this last limitation, preclinical experiments in which Treg specificity is engineered with transgenic TCRs or chimeric antigen receptors (CARs) revealed that, unlike polyclonal Treg, antigen-directed Treg were highly effective in halting and reversing autoimmunity (9–12). While methods for adoptive transfer of Treg with engineered specificity are currently under clinical development (13), pharmacological approaches to enhance endogenous antigen-specific Treg are potentially more scalable and warrant further investigation.

In a wide range of mouse models of autoimmunity, allergy, and transplant rejection, Treg enrichment with IL-2/antibody complexes (IL-2c) (14) prior to or during antigen challenge consistently suppressed disease and pathology (15–21), raising the possibility that antigen priming in the presence of elevated Treg frequencies will durably alter the balance between antigen-specific Treg and pathogenic Tconv in patients. An example of combined Treg enrichment and ASI in mice is the complete suppression of factor VIII antibody formation in FVIII-deficient mice treated with a combination of IL-2c and FVIII protein (22, 23). In this setting, FVIII-specific Treg were elicited and, upon isolation and transfer, could establish tolerance in untreated mice. These studies also demonstrated that Treg enrichment before FVIII challenge was more effective than concomitant treatment (23). Leveraging the same system, we recently reported that combining a Treg-selective Fc-fused IL-2 mutein (Fc.IL2m) with FVIII gene therapy also established durable suppression of FVIII antibody formation (24, 25).

To bridge these findings to clinical trial design, more research is needed to find the optimal dosing strategies that enforce tolerogenic responses to ASI and increase the relative abundance of antigen-specific Treg, and our goal in this work was to define a specific dosing regimen that permanently augmented this ratio. Because antigen-stimulated Tconv should upregulate CD25 and consequently increase Fc.IL2m-responsiveness, we considered the timing of ASI following Fc.IL2m treatment to be an important parameter to optimize to ensure that residual Fc.IL2m would not act on ASI-stimulated CD25⁺ Tconv. Our characterization of Fc.IL2m molecules suggested that the timepoint of peak Treg enrichment was a likely indicator of the moment when Fc.IL2m fell below biologically active levels (25); thus, we reasoned that ASI at timepoints centered around peak Treg enrichment should reveal the best strategy. With an optimal dosing strategy selected, we then explored the durability of antigen-specific Treg enrichment following treatment cessation and through acute viral infection.

MATERIALS AND METHODS

Mice

Female 5–6 week old C57BL/6J (Stock No: 000664) and B6.SJL-Ptprc^a Pepc^b/BoyJ (Stock No: 002014) (CD45.1) mice were purchased from the Jackson Laboratory (Bar Harbor, ME). Female OT-II mice were bred in house, and maintained in specific pathogen-free conditions. All experiments involved female mice at the age of 8–13 weeks. Experiments were executed under an animal study proposal approved by the Animal Care and Use Committee (ACUC) of Benaroya Research Institute (Seattle, WA).

Adoptive transfer

Cell suspensions were generated from pooled inguinal, mesenteric, brachial, axillary, and superficial cervical lymph nodes isolated from female OT-II mice. Two million lymph nodes cells in 50 μL PBS were injected retro-orbitally (R.O.) into anaesthetized B6.SJL mice. Anesthesia was induced by 1–4% Isoflurane.

Dosing regimens of Fc.IL2m and DEC-OVA

For testing the pharmacodynamics of Treg enrichment, C57BL/6J mice (n=5) received a single 3 μg intraperitoneal dose of Fc.IL2mut27 (Fc.IL2m), and flow cytometry was performed with saphenous vein blood samples obtained on days 0, 3, 4, 5, and 7.

Four days after the adoptive transfer of OT-II lymph node cells, recipient B6.SJL mice were injected intraperitoneally with 3 μg Fc.IL2m or PBS, after which mice received 0.5 μg anti-DEC205-OVA or PBS on day 2, 3, or 4 after Fc.IL2m treatment. Mice received 1 or 3 weekly Fc.IL2m/DEC-OVA treatment cycles. After the last treatment cycle, mice were euthanized, and lymph nodes and spleens were harvested for immunophenotyping.

Due to the large size of these experiments, not all treatment groups were included in each replica experiment. Figures represent the aggregate of all replica experiments, and the number of individual experiments performed is noted in the figure legends.

VSV-OVA infection

Following 3 Fc.IL2m/DEC-OVA treatment cycles (D22), B6.SJL hosts were infected intranasally with vesicular stomatitis virus encoding ovalbumin (VSV-OVA) at a dose 10⁴ PFU in 50 μL in sterile PBS. For this procedure mice were anaesthetized with 1–4% Isoflurane. Infected mice were housed in an ABSL-2 room.

Enrichment of OT-II donor T cells

Pooled spleen and lymph node cells were isolated from B6.SJL recipient mice and red blood cells were lysed with Gibco ACK lysing buffer at room temperature for 5 minutes. Washed cell suspensions were counted (Invitrogen Countess II automated cell counter) and a small aliquot was retained for immunophenotyping, and remaining cells were pelleted and resuspended in 45 μL MACS buffer per 10⁷ cells. Anti-CD45.2-PE (5 μL per 10⁷ cells) was added to each sample and incubated for 30 minutes at 4°C. Cells were washed and resuspended in 90 μL of MACS buffer per 10⁷ cells, and were incubated with anti-PE microbeads (Miltenyi, Auburn, CA) at 20 μL per 10⁷ cells for 10 minutes at 4°C. After washing and resuspension in 500 μL of MACS buffer per 10⁸ total cell numbers, cells were applied to LS MACS columns in the magnetic column holder (Miltenyi, Auburn, CA). Columns were washed three times with 3 mL of MACS buffer, removed from the magnetic separator and placed on 15 mL collection tubes. MACS buffer (5 mL) was added to each column and cells were flushed out by firmly pushing the plunger into the column. Column-bound cells were counted and stained for flow cytometric analysis.

Flow cytometry and statistics

For surface staining, cells were incubated at 4°C for 30 minutes in staining buffer (0.5% BSA in PBS) with the following conjugated antibodies (from Biolegend unless otherwise specified): anti-CD3 (145-2C11, BD), -CD4 (RM4–5), -CD8a (53-6.7), -CD25 (PC6.1), NK1.1 (PK136), -CD62L (MEL-14), -CD44 (IM7), -CD45.2 (104), -CD45.1 (A20), TCR-Vβ5.1/5.2 (MR9–4), and TCR-Vα2 (B20.1). All samples were also stained for dead cells with Zombie Aqua kits from Biolegend according to the manufacture instructions. Cells were washed after surface staining and permeabilized for 30 minutes with eBiosience Fix/Perm buffer at 4°C. In PermWash staining buffer, cells were stained for 30 minutes at 4°C with anti-Foxp3, (FJK-16S, eBioscience), -Ki67 (B56, BD), and -Helios (22F6, Invitrogen). To allow for OT-II T cell enumeration, the entire column bound fraction was analyzed by flow cytometry. Data were acquired on LSRII flow cytometer (BD Bioscience), and analyzed using FlowJo software (Treestar) (Supplemental Fig. 1). Only mice yielding > 50 detectable donor OT-II cells were included in the analyses and figures. Student’s T-tests were used to compare population pairs. *, p<0.05; **, p<0.01; ***, p<0.001; ****, p<0.0001; *****, p<0.00001.

RESULTS

Expansion of antigen-specific Treg following repeated cycles of ASI during the peak of Fc.IL2m-mediated Treg enrichment

To track the frequency of antigen-specific Treg in mice with a polyclonal T cell repertoire, we utilized congenically marked B6.SJL mice that received 2×10⁶ LN cells from OT-II TCR transgenic mice, which, owing to endogenous TCR expression, contain OVA-reactive Treg at a frequency of ~3–6% of CD4⁺ T cells (26, 27). In this system, the frequency of OT-II T cells in recipient mice was only ~10-fold higher than the typical frequency of naïve CD4 T cells specific for a single peptide antigen (28), necessitating donor cell enrichment by magnetic bead selection to enumerate and phenotype rare OT-II T cells after the treatment regimen (29). For the ASI modality, we employed the anti-DEC205-OVA fusion protein (DEC-OVA), which targets OVA to DEC205-expressing dendritic cells (30). We selected DEC-OVA as our ASI because previous reports have shown that antigen delivered to DEC205⁺ DCs can stimulate both proliferation of naturally occurring Treg (nTreg) and formation of new Treg from peripheral Tconv (pTreg), potentially increasing the likelihood of achieving antigen-specific Treg enrichment with our protocol (31–33). We selected a 0.5 μg dose of DEC-OVA as this dose should be processed and presented by MHC class II on DC within 3 hrs and stimulate OT-II cells for up to 2–3 days following a single injection (30). In our initial evaluation of Fc.muteins, Fc.Mut27 (IL-2 mutations M33D, V106D) and the subsequently well-characterized Fc.Mut24 (N103R, V106D) (25) possessed very similar CD25-dependency, in vivo half-life, and Treg-selectivity (Supplemental Fig. 2); and Fc.Mut27 (hereafter referred to as Fc.IL2m) was used for this work.

Following a single injection of 3 μg Fc.IL2m, Treg in blood increased 3-fold on day 3, returning to baseline by day 7 (Fig. 1A). Because Fc.IL2m treatment of OT-II recipient mice should increase the frequency of both OT-II and host (B6.SJL) Treg, subsequent immunization with DEC-OVA during peak Treg enrichment should further stimulate OT-II Treg growth and survival, and confer a numerical advantage over other Treg following their return to baseline frequency on day 7. OT-II Tconv responses to DEC-OVA in the presence of expanded OT-II and host Treg may also enable or enhance their conversion into pTreg through processes akin to infection tolerance (34). Thus, we set out to determine the optimal day to administer DEC-OVA following each weekly Fc.IL2m dose. DEC-OVA dosing was evaluated shortly before, during, and after peak Treg frequencies, or day 2, 3, and 4 after each Fc.IL2m dose. Three treatment cycles were performed, as we postulated that repeated Fc.IL2m/DEC-OVA treatment would elicit more robust OT-II Treg enrichment. The treatment regimen was initiated four days following OT-II cell transfer, and on day 22, CD45.2⁺ donor cells were column-enriched from splenocyte and lymph node cell suspensions and OT-II Treg and Tconv were enumerated by flow cytometry (Fig. 1B).

FIGURE 1. Repeated cycles of ASI during peak Fc.IL2m-mediated Treg enrichment results in sustained expansion of antigen-specific Treg.

(A) A single 3 μg dose of Fc.IL2m results in maximal Treg enrichment in peripheral blood on day 3, which returns to baseline on day 7. (B) Schematic of exploratory Fc.IL2m/ASI treatment regimen in B6.SJL recipient mice harboring OT-II CD4 T cells. (C) Representative flow cytometry plots and aggregate data showing the percentage of Treg among CD45.2 OT-II CD4 T cells. (D) Percent Treg among total CD4 T cells in recipient mice. (E) Frequency of OT-II Tconv (CD4+ Foxp3⁻) cells among total recipient CD4 T cells. Open diamonds: Vehicle (PBS); filled circles: Fc.IL2m.

Data represent the aggregate of 4 individual experiments.

Consistent with our initial hypothesis, we found that the Fc.IL2m/ASI combination treatment was most effective at enriching OT-II Treg when DEC-OVA was administered on day 2 or 3 after each Fc.mutein dose (Fig. 1C). In contrast, treatment with DEC-OVA alone modestly increased the frequency of OT-II Treg, and treatment with Fc.IL2m alone had no lasting effect. The durable nature of Treg enrichment with the Fc.IL2m/DEC-OVA combination treatment regimen was specific to OT-II cells, as the frequency of Treg among total host CD4 T cells returned to baseline at day 22 (Fig. 1D). Importantly, the combination treatment did not alter the frequency of OT-II Tconv among total CD4 T cells, and a significant negative correlation between the % Treg among OT-II CD4 T cells and overall OT-II Tconv frequency was not observed, indicating that OT-II Treg enrichment resulted from OT-II Treg expansion rather than OT-II Tconv contraction (Fig. 1E, Supplemental Fig. 3). Thus, the combination of Treg enrichment and ASI imprinted an elevated Treg frequency among antigen-specific CD4 T cells. For subsequent experiments, we chose to administer DEC-OVA on day 3 following Fc.IL2m or vehicle dosing, as this timepoint resulted in the most significant increase in OT-II Treg frequency relative to Fc.IL2m treatment alone.

Antigen-specific Treg enrichment requires multiple Fc.IL2m/ASI cycles and persists following treatment cessation

For our protocol to serve as an effective Treg vaccine, it was important to characterize the durability of antigen-specific Treg enrichment following treatment cessation. To explore this, we performed a similar set of experiments in which OT-II recipient mice were treated with weekly DEC-OVA alone, or DEC-OVA 3 days after individual Fc.IL2m doses. Following 3 treatment cycles, some mice were evaluated at the completion of the 3^rd cycle, while others were left untreated for an additional 22 days. Furthermore, to determine whether 3 cycles were indeed necessary to observe any OT-II Treg enrichment, we included a third set of mice that received only 1 treatment cycle (Fig. 2A).

FIGURE 2. More than one Fc.IL2m/ASI cycle is necessary for antigen-specific Treg enrichment, which persists for at least 3 weeks following treatment cessation.

(A) Schematic of comparison of 1 and 3 weekly treatment cycles, and 3 cycles followed by 3 weeks without treatment. (B) Percentage of Treg among OT-II CD4 T cells in different treatment groups. (C) Frequency of OT-II Tconv among total recipient CD4 T cells.

Data represent the aggregate of 2 individual experiments.

In comparing 1 and 3 weekly cycles of Fc.IL2m/DEC-OVA, DEC-OVA alone, or vehicle control groups, we found no evidence of enrichment among mice receiving only 1 cycle, while 3 cycles of the combination treatment again elicited more than 3-fold OT-II Treg enrichment (Fig. 2B). Notably, this expansion persisted for an additional 3 weeks following treatment cessation (Fig. 2B). As seen in our previous experiment, the frequency of OT-II Tconv did not differ between groups (Fig. 2C). Thus, the Fc.IL2m/ASI combination protocol was capable of establishing a durable increase in antigen-specific Treg.

Regarding Treg phenotype following 3 treatment cycles, expression of transcription factor Helios was similar among the different treatment groups, and was slightly lower in OT-II Treg relative to host Treg (Supplemental Fig. 4). CD25 and Ki67 expression was also similar among treatment groups, with a slight increase in Ki67⁺ cells in the Fc.IL2m/ASI combination group (Supplemental Fig. 4), potentially reflecting increased proliferation in response to both stimuli. Relative to host Treg, overall CD25 levels were slightly higher in OT-II Treg regardless of treatment group, a phenotype that may have resulted from stimuli present in the niche populated by transferred OT-II Treg (Supplemental Fig. 4).

Expanded antigen-specific Treg persist through an acute anti-viral immune response directed against the tolerizing antigen

We next asked whether the Fc.IL2m/DEC-OVA-exposed OT-II population would retain its high Treg frequency following antigen stimulation in a proinflammatory context, and whether the enriched OT-II Treg would suppress the OT-II Tconv response to the same challenge. Following OT-II transfer and 3 weekly cycles of the different treatment groups, we intranasally infected half of the mice in each group with OVA-expressing vesicular stomatitis virus (VSV-OVA) (day 22) and analyzed column enriched OT-II cells 1 week post infection (day 29) (Fig. 3A).

Fig. 3. Phenotypes of OT-II and host Treg and Tconv following 3 treatment cycles (day 22 of Figure 2A).

(A) Examples of gate placement for OT-II Treg. The same gates were applied for other populations. (B) Helios and CD25 geometric MFI (top) or percentage of Helios⁺, CD25⁺, or Ki67⁺ cells (bottom) for OT-II Treg of the 3 treatment groups and host Treg for all treatment groups combined. Flow cytometry data containing more than 10 Treg cells were included in the analysis. (C) Percentage of CD25⁺ or Ki67⁺ among OT-II Tconv in the 3 treatment groups and host Tconv for all treatment groups combined.

All host T cell data was derived from the pooled spleen + LN cell suspension prior to column enrichment of CD45.2⁺ T cells.

Following intranasal VSV-OVA infection, OT-II cells adopted a CD44⁺ CD62L¯ effector T cell phenotype, indicating robust T cell activation (Fig. 3B). Among the uninfected animals, the Fc.IL2m/DEC-OVA combination treatment increased OT-II Treg 4.5-fold, while less robust increases were observed in the Fc.IL2m and DEC-OVA groups (Fig. 3C). Following VSV-OVA infection, however, OT-II Treg frequency decreased in the PBS and Fc.IL2m groups, yet remained high in the DEC-OVA treated groups. Notably, among VSV-OVA infected mice, the combination treatment resulted in an 11- and 14-fold increase in Treg percentages relative to the vehicle and Fc.IL2m groups, respectively (Fig. 3C, right). As in previous experiments, no significant change in OT-II Tconv frequency was observed among the groups, except for modest increases following VSV-OVA infection in the vehicle and Fc.IL2m groups and modest decreases in the DEC-OVA and combination treatment groups, with the most significant difference being a 5-fold decrease in the Fc.IL2m/ASI group following infection (Fig. 3D). Consistent with this observation, VSV-OVA-induced Ki67 in OT-II Tconv was reduced in DEC-OVA treated mice, and significantly so in mice co-treated with Fc.IL2m (Fig. 3E). VSV-OVA infection also induced proliferative responses in OT-II Treg, which were partially attenuated by prior DEC-OVA exposure (Supplemental Fig. 4). This finding suggests that previous DEC-OVA exposure rendered the OT-II cells resistant to infection-induced expansion, either due to cell intrinsic processes such as anergy or due to dominant suppression mediated by the enriched OT-II Treg. Thus, the Fc.IL2m/ASI treatment protocol is capable of durably expanding antigen-specific Treg that persist through inflammatory processes and may suppress Tconv reactive to the same antigen during acute viral infection.

DISCUSSION

The goal of ASI is to re-establish immune tolerance in autoimmune patients by delivering self-antigens or allergens to tolerize rather than activate B and T lymphocytes, and, as such, is considered a tolerizing vaccination strategy (35–37). In laboratory mice, which harbor a uniquely quiescent immune system, tolerogenic vaccination is effective when antigen is simply delivered without additional response-modifying agents (38). These tolerizing approaches include vaccination with 1) free peptide epitopes, 2) nanoparticles bearing peptide, whole antigen, or MHC-peptide complexes, 3) antigen bound to whole cells or RBCs, and 4) antigen coupled to antibodies that target dendritic cells—as we have employed here (30, 35, 38, 39). In humans, and perhaps more so in autoimmune or allergic patients, chronic immune homeostasis exists in a more activated and variable state than that of laboratory mice, making it difficult to predict how an individual will respond to antigen stimulation of pro-inflammatory lymphocytes in the absence of immunomodulators. Consequently, multiple groups are exploring immunosuppressive adjuvants to enhance tolerogenic responses. These approaches are largely focused on specific molecular entities, such as inhibitors of immunostimulatory pathways (e.g. abatacept or rapamycin), and agonists of inhibitory pathways (recombinant TGF-β or IL-10) (37, 39). It may be difficult, however, to determine which pathway will be most active in particular patient groups and whether manipulating one pathway will elicit compensatory mechanisms that diminish efficacy.

In this work, we have taken a cellular approach—Fc.IL2m-mediated Treg enrichment—to enforce tolerogenic responses to ASI. Treg employ multiple adaptive suppressive pathways to restrain different types of inflammation, increasing the likelihood that Treg enrichment will reliably control responses to ASI in different diseases and patient populations (6). For our experimental system, we utilized recipient mice of OT-II TCR transgenic T cells, which contain a low proportion of OVA-responsive Treg. Transferred OT-II T cells were present at frequencies of 10⁻⁵ to 10⁻⁴, retained their baseline percentage of Treg, and could be recovered and enumerated by magnetic bead selection and flow cytometry.

Consistent with our hypothesis that ASI administration at the peak of Treg enrichment would be most effective, we found that DEC-OVA dosing at day 2 and day 3 after Fc.IL2m treatment were roughly equivalent at maintaining higher OT-II Treg frequencies while day 4 ASI was less effective. Interestingly, 1 cycle was not sufficient to increase OT-II Treg, while 3 cycles resulted in a significant 3- to 4-fold OT-II Treg enrichment. Thus, Treg accumulation may have resulted from an incremental step-wise process in which each cycle preferentially activated proliferative and survival pathways in OT-II Treg, and it is interesting to speculate that additional cycles would further increase OT-II Treg frequencies. Importantly, OT-II Treg enrichment persisted for at least 3 weeks following treatment cessation and persisted through a strong, acute inflammatory response to VSV-OVA infection. In the vehicle and Fc.IL2m-alone groups, VSV-OVA infection led to a marked decline in OT-II Treg, a finding that is consistent with previous research on the fate of naturally occurring virus antigen-responsive Treg following infection (40). In contrast, OT-II Treg percentages remained similar following VSV-OVA infection in DEC-OVA groups with the highest frequency being in mice that also received Fc.IL2m. Thus, while Fc.IL2m alone can promote Treg expansion, TCR stimulation is required for maintaining Treg numbers or competitive fitness during an anti-viral response. This study demonstrates the durability of the Fc.IL2m/ASI approach and suggests that a relative loss of Treg among CD4 T cells promoting inflammatory processes could be reversed by this protocol.

It should be noted that mice treated with anti-DEC205-fused antigen and other sub-immunogenic methods of antigen delivery can develop pTreg that co-express Helios, a transcription factor that is also expressed in thymically-derived or naturally occurring Treg (tTreg and nTreg) and that enforces Treg phenotypic stability (31–33, 41–44). While we cannot rule out that some Treg expansion resulted from de novo pTreg formation, we are unaware of reports of pTreg development from OT-II Tconv (e.g., from RAG-deficient OT-II T cells) following OVA administration other than oral delivery (45), and it has been postulated that OT-II Tconv have a poor intrinsic capacity to convert into pTreg (46). Regardless of their origin, the similarity in Helios expression between host Treg and OT-II Treg of the different treatment groups suggests the expanded Treg should possess the phenotypic stability of nTreg (47), consistent with our observed maintenance of OT-II Treg enrichment following VSV-OVA infection. Defining the transcriptome and epigenome of expanded antigen-specific Treg is warranted to further characterize their phenotypes and functional properties.

Subsequent evaluation of this approach should explore different dosing strategies and efficacy in suppressing autoimmunity. In mice, higher Fc.mutein doses and longer treatment cycles should be tested (e.g. 10–15 μg of Fc.Mut24 (25) with 10–14 day cycles), and efficacy could be evaluated in spontaneous autoimmune models such as (NZB × NZW)F1 lupus with the Hsp70 autoantigen (48) or NOD T1D with pancreatic islet antigens. In patients, similar approaches could be attempted with half-life-extended IL-2 muteins and ASI modalities currently in clinical development, where antigen-specific Treg enrichment could be monitored with existing MHC tetramer technologies (40). Treatment timepoints should be adjusted based on the pharmacodynamics of Treg enrichment and the pharmacokinetics of the IL-2-based therapeutic to ensure that ASI is administered after the IL-2 therapeutic falls below levels that will stimulate ASI-induced CD25⁺ Tconv. Furthermore, cycle duration should be extended based on ASI pharmacodynamics to increase the likelihood that ASI-induced CD25 expression on responding Tconv declines before subsequent IL-2 therapy doses. Indeed, stimulating Fc.IL2m-expanded Treg with ASI may be necessary to fully realize their potential to establish lasting tolerance and may obviate a requirement for sustained Treg enrichment through chronic IL-2 mutein treatment.

FIGURE 4. OT-II Treg enrichment persists through acute VSV-OVA infection.

(A) Schematic of experimental plan, in which OT-II recipient mice received 3 treatment cycles followed by intranasal VSV-OVA infection. (B) Percentage of CD44⁺ CD62L⁻ OT-II Tconv. (C) Percentage of Treg among OT-II CD4 T cells in different treatment groups. (D) Frequency of OT-II Tconv among total recipient CD4 T cells. (E) Percentage of Ki67⁺ cells among OT-II Tconv.

Data represent the aggregate of 2 individual experiments.