Hybrid cytokine IL233 renders protection in murine acute graft vs host disease (aGVHD)

Abstract

Previously, we generated IL233 - a hybrid cytokine composed of interleukin (IL)-2 and IL-33, with better therapeutic potential than either cytokine in multiple inflammatory diseases, in part through promoting T-regulatory cells (Tregs). Here we test the potential of IL233 pretreatment in a murine model of excessive Th1 activation, the parent-into-F1 model of acute GVHD (aGVHD). Five days of IL233 pretreatment of the recipients blocked or delayed the aGVHD-linked loss of B-cells as seen in either the peripheral blood (day-11) or lymph nodes (day-14). IL233 pretreatment also prevented the expansion of donor CD8 T-cells in blood and LN at day-14 and significantly reduced day-14 serum IFNγ and TNFγ compared to saline treated GVHD mice although, the level of Tregs did not statistically differ between saline and IL233-treated mice. Overall, the current study provides support for the use of IL233 as a therapeutic option in excessive Th1/CD8-driven conditions.

1. Introduction

IL233 is a recently generated hybrid cytokine that combines the potential of IL-2 and IL-33 as a single molecule, where mature IL-2 (Ala₂₁-Gln₁₆₉) is fused to the amino-terminus of mature IL-33 (Ser₁₀₉-Ile₂₆₆) via a flexible 15 amino-acid (GlyGlyGlyGlySer)₃ linker [1]. When administered in vivo, IL233 results in a rapid, robust and more durable expansion of Tregs than when either IL-2 or IL-33 were administered alone or as a mixture [1, 2]. The premise being that IL233 enforces the co-localization of both IL-2 and IL-33 activities to utilize independent, yet complimentary pathways for improved Treg homeostasis. IL233 administration has shown an ability to ameliorate disease in preclinical models of ischemic and nephrotoxic acute kidney injury, type-2 diabetes and diabetic nephropathy as well as autoimmune diseases including lupus glomerulonephritis [1–4]. The protection observed by IL233 in these preclinical studies was due in-part to an increase in Tregs because protection was lost with depletion of Tregs with anti-CD25 antibody [3]. The observed effects of IL233 in our previous studies were also ST2-dependent, as observed by increased production of IL-5 and IL-13 and increased recruitment of eosinophils and M2 macrophages in the adipose tissue in a murine model of type-2 diabetic nephropathy [4]. IL233 treatment also inhibited the activation and proliferation of Th1 cells as evidenced with a strong decline in the proportion of IFNγ and TNFγ producing T-cells [1–3]. These studies raise the possibility that IL233 could be beneficial in other conditions characterized by excessive or unwanted Th1 cell activation and ineffective downregulation.

The parent-into-F1 (P→F1) model of acute GVHD is a useful pre-clinical model of in vivo Th1 immune responsiveness (reviewed in [5]). Disease is induced by the transfer of homozygous parental strain T cells (both CD4 and CD8) into normal (unirradiated) semi-allogeneic F1 hosts. Because the F1 is tolerant to parental strain MHC molecules, no irradiation or conditioning of the host is required thus minimizing the non-specific cytokine storm typically observed in bone marrow transplant models. Disease is a consequence of a strong in vivo alloantigen driven response and is initiated by donor CD4 T cells recognition of F1 allogeneic MHC II leading to IL-2 production and proliferation [6]. Donor CD8 T cells recognition of F1 allogeneic MHC I in conjunction of donor CD4 T cell help (IL-2) results in maturation of effector CD8 CTL that then eliminate host lymphocytes such that by day 14 after donor cell transfer, F1 splenic B cells are significantly depleted [5]. Advantages of this model are its relatively brief turn-around time, the identity of the initiating antigens is known and importantly, disease severity can be modulated up or down by increasing or decreasing respectively the number of donor cells transferred. Using this model, we sought to determine whether pretreatment with IL233 could lessen acute GVHD severity. Our results add further support to the use of IL233 in conditions where immune response downregulation is desired.

2. Materials and Methods

2.1. Animals and treatments

All procedures were approved by the University of Virginia Institutional Animal Care and Use Committee and in accordance with the National Institutes of Health (NIH) Guide for the Care and Use of Laboratory Animals. Female B6D2F1 and C57BL/6J (B6) mice were purchased from the Jackson laboratories (approximately 10 weeks old). Recombinant IL233 hybrid cytokine was produced in E.coli using methods described earlier [1]. Cohorts of B6D2F1 recipient mice were injected (i.p.) with saline or IL233 (66pmoles) daily for 5 days before receiving 40 x 10⁶ splenocytes (i.v.) from C57BL/6J (B6) mice at day 8. After splenocyte transfer, peripheral blood samples were taken from all groups at day 4, 8, 11 and 14. Mice were euthanized at day 14 and samples harvested and analyzed.

2.2. Flow Cytometry

Flow cytometry was performed as described previously [1]. Antibodies used for analysis are listed below. Data were acquired on a FACScan cytometer (BD Biosciences) with a 5-color upgrade (Cytek Development) and analyzed with FlowJo™ software (BD Biosciences) as described in the results section. The donor cells were gated as H-2Kb⁺H-2Kd⁻ cells and the recipient cells were gated as H-2Kb⁺H-2Kd⁺ cells. Fluorescently labeled anti-mouse antibodies against the following molecules with the respective clones indicated in parenthesis were from BioLegend unless otherwise indicated: CD4 (GK5.1), CD8α (53–6.7), TCRβ (H57–597), B220 (RA3–6B2), CD44 (IM7), CD62L (MEL-14), IFNγ (H22), TNFα (MP6-XT22), H-2K^b (AF6–88.5), H-2K^d (SF1–1.1) and Foxp3 (FJK-16, Life Technologies). The gating strategy employed in the current study is presented in supplementary figure 1.

2.3. Cytokine analysis

For analysis of intracellular cytokine production, single-cell suspension from pooled peripheral lymph nodes were stimulated for 5 hours with phorbol 12-myristate 13- acetate (PMA, 20 ng/ml) and ionomycin (1g/ml) in the presence of 1mM monensin (all from Sigma, USA). Cells were stained with antibodies for CD4, CD8, IFN? and TNFa using the intra-cellular staining kit (Life Technologies). Recipient CD4 and CD8 gated cells (H-2Kb⁺H-2Kd⁺) were analyzed for intracellular cytokines by flow-cytometry as before [1]. The serum cytokines (IFNγ, TNFα, IL-4 and IL-10) were analyzed using ELISA-MAX™ kits, using the manufacturer’s instructions (BioLegend).

3. Result and discussion

3.1. IL233 pretreatment blocks B cell elimination in aGVHD mice.

The B6→B6D2F1 murine model of aGVHD has been successfully used to investigate mechanisms of in vivo CD8 CTL generation [5]. Donor strain CD8 CTL specific for F1 allogeneic MHC have been demonstrated by in vitro, ex vivo and in vivo assays [5, 7, 8]. Additionally, a reduction in the numbers of F1 splenic B cells has been shown to be a highly sensitive surrogate marker of in vivo parental CD8 CTL activity [7]. To determine whether in vivo IL233 is able to mitigate aGVHD, we chose to administer IL233 as a 5-day pretreatment. This 5-day pretreatment strategy was opted based on information available on the successful use of low dose IL-2 for immune modulation in type 1 diabetes [9–11] and our successful use in other model systems [1–4]. Additional preclinical work is ongoing with IL233 for a therapeutic approach, when IL233 will be administered after the induction of GVHD. Because the optimal timing of IL233 administration in aGVHD is not known, we induced a mild aGVHD to better detect subtle IL233 effects. F1 mice received B6 donor cell splenocytes just at the threshold of disease induction and were sequentially monitored over the time course of the experiment (0–14 days). Flow cytometry was used to quantitate the number of B220⁺ lymphocytes in peripheral blood rather than spleen so as to determine sequential changes in individual mice. As shown in Figure 1A, peripheral blood B cells started to decline in saline treated aGVHD mice as early as day 4 and continued to decline until day 14 consistent with aGVHD phenotype. Pretreament of aGVHD mice with IL233 resulted in a significant attenuation of aGVHD mediated B cell elimination through day 11 however by day 14, they had declined to levels not significantly different vs. saline-control aGVHD mice. Examination of day 14 LN revealed that IL233 significantly increased host B cells in aGVHD mice as compared to saline control GVHD mice measured either as a percentage (Fig. 1B) or as total numbers (Fig. 1C) of host B220+ lymphocytes. Of note, saline treated aGVHD mice also showed a significant reduction in B cell percentage in peripheral lymph nodes at day 14 vs. either uninjected normal F1 mice or IL233 treated aGVHD mice. Total B cell numbers in saline treated control aGVHD mice were not significantly reduced vs. normal F1 LN likely reflecting the mild aGVHD that was induced. Taken together, these results indicate that IL233 pretreatment in mild aGVHD mice is able to mitigate the loss of F1 B cells characteristic of aGVHD.

Figure 1. IL233 pretreatment prevents aGVHD induced loss of B cells.

Approximately 10-week old female B6D2F1 mice were treated with saline or IL233 for 5 days, followed by infusion (i.v.) of 40 x 106 splenocytes prepared from 10–12 weeks old female B6 mice. A) Peripheral blood samples were obtained from control, saline and IL233 pre-treated mice at day 4, 8 and 11 and the number of recipient F1 B lymphocytes (B220⁺ cells) quantified by flow cytometry. B-C) Pooled peripheral lymph nodes (cervical, axillary, brachial and inguinal lymph nodes) were obtained at day 14 (terminal endpoint) and the number of F1 B lymphocytes measured by flow cytometry. F1 B lymphocytes are represented as percentage changes in (B) and as absolute cell numbers in (C). Data are shown as mean ± S.E.M and are representative of 6–10 mice per group. *p < 0.05 between control and saline; # p < 0.05 between control and IL233; $ p < 0.05 between saline and IL233 as calculated by multiple comparison using one-way ANOVA.

3.2. IL233 treatment impairs donor CD8 T cell expansion and reciprocal host-vs-graft (HVG) response likely through Treg independent mechanisms.

The elimination of splenic F1 B cells that is characteristic of aGVHD in the P→F1 model is mediated by donor strain CD8 CTL specific of F1 allogeneic MHC I [5]. Typically, the numbers of donor effector CD8 CTL in the spleen peak at days 10–12 after donor transfer [7]. The kinetics of engrafted donor CD8 T cell in PBL has not been previously characterized. As shown in Figure 2A, donor CD8 T cells are detected in the peripheral blood by day 4 for both saline treated and IL233 treated aGVHD mice, however from days 11–14 the donor CD8 T cell numbers decline significantly only for IL233 treated aGVHD mice. Similarly, a significant reduction in donor CD8 T cells in IL233 treated vs. saline treated aGVHD mice was observed in lymph nodes at day 14 (Figure 2B). Together, these results support the idea that the impaired elimination of F1 B cells mediated by IL233 seen in Figure 1 is a consequence of reduced numbers of donor CD8 T cells and with it a reduction in effector CD8 CTL capable of eliminating host B cells.

Figure 2. IL233 attenuates aGVHD by preventing expansion of CD8 T cells.

Using the same experimental protocol as Fig. 1, the numbers of donor CD8⁺ T (H-2Kb⁺H-2Kd^- cells), expressed as fold change over saline group, were measured sequentially by flow cytometry in the peripheral blood obtained at the days indicated (A); and in the pooled lymph nodes (LN) at day 14 (B). Similarly, recipient (H-2Kb⁺H-2Kd⁺ cells) CD8⁺ T cells (% of lymphocytes) were measured sequentially in peripheral blood (C) and at day 14 in pooled LN (D). The ratio of recipient CD8 T effector memory:T central memory at day 14 is shown for: peripheral blood (E) and LN (F). Sequential changes in peripheral blood regulatory T cells (CD4⁺Foxp3⁺ cells) are shown in (G) and in LN at day 14 as percentage changes in (H) and as absolute cell numbers in (I). Data are shown as mean ± S.E.M and are representative of 6–10 mice per group. * p < 0.05 between control and saline; ^# p < 0.05 between control and IL233; ^$ p < 0.05 between saline and IL233 as calculated by multiple comparison using one-way ANOVA and unpaired t-test.

Previous work has shown that P→F1 aGVHD mice exhibit a reciprocal increase in host splenic CD8 T cells at about Day 12–14 as part of a host-vs.-graft (HVG) response, the strength of which parallels the degree of GVHD response [7]. We observed a significant increase in host CD8 T cells in saline treated control aGVHD mice for PBL at days 11–14 (Figure 2C) and peripheral lymph nodes at day 14 (Figure 2D) as compared to uninjected F1 mice consistent with the expected HVG response. By contrast, IL233-treated aGVHD mice did not exhibit a significant increase in CD8 T cells in either PBL or lymph nodes vs. uninjected control F1 mice (Figures 2C, 2D). These results are consistent with the reduced donor CD8 T cell numbers in IL233 treated mice (Figure 2A) that in turn induces a reduced HVG response.

Since the balance between T central memory (T_cm) and T effector memory (T_em) cells is a predictor of immune activation [12], and that the differential expansion of T_cm and T_em is an indicator of antigen-specific expansion [13], we calculated the ratio of T_em to T_cm (T_em/T_cm) by flow cytometry. As shown in Figure 2E–2F, saline treated mice had a more pronounced CD8 T_em shift in peripheral blood and lymph nodes compared to controls. IL233-treated mice had a trend for attenuated CD8 T_em response in blood and showed a significant attenuation in lymph nodes with the T_em/T_cm ratios being similar to control mice indicating a reduction in T cell activation and severity of aGVHD. No differences in CD4 T cell memory status was observed in recipient animals (data not shown).

We recently reported protective effects of IL233 in preventing acute kidney injury, nephrotoxicity, lupus glomerulonephritis and type 2 diabetic nephropathy[1–4]. In those studies, IL233 administration led to a robust expansion of Tregs along with the observed protective effects that were lost upon Treg depletion using anti-CD25 mAb. We therefore investigated the effects of IL233 on the levels of Foxp3⁺CD4⁺ Treg cells in acute GVHD. Peripheral blood samples in mice treated with IL233 showed elevated levels of Tregs as early as day 4 in both saline and IL233 -treated mice compared to controls, however, the levels declined by day 8 and stayed similar to control mice thereafter (Figure 2G). Both saline and IL233 -treated mice exhibited significantly elevated levels of Tregs in lymph node samples at day 14 compared to untreated control mice with IL233 treatment only showing a higher statistically non-significant trend over the saline-treated group that was a little more pronounced in absolute count (Figure 2H–I). These results contrast with our earlier studies, where we observed robust increase in Tregs upon IL233 compared to the saline control treated nephropathy [1–3], suggesting that: a) IL233 pretreatment may utilize mechanisms other than Treg-proliferation to offer protection from aGVHD: or b) that any Treg boosting effect of IL233 pretreatment had worn off by day 4. Additionally, one of the mechanisms identified in our previous studies is that Tregs isolated from IL233-treated mice had higher suppressive activity than those isolated from the saline treated mice [1, 2], which could be operative in these experiments. Another possibility for lack of significant difference in Tregs between the saline and IL233 groups could be due to an expansion of Tregs populations due to the allo-response induced IL-2 production masking the IL233 effect [14]. This is evidenced by a higher level of Tregs in the saline group as compared to the control group (Figure 2G–I).

3.3. Treatment with IL233 inhibits proinflammatory cytokine production in aGVHD

In the P→F1 model, splenic IFNγ and TNFα elevation are reliable markers for aGVHD phenotype [8, 15, 16] and TNFα blockade inhibits aGVHD [16]. We therefore examined the levels of IFNγ and TNFα and other cytokines to confirm whether IL233 mitigates these features of aGVHD. As shown in Figure 3A, levels of IFNγ and TNFα were significantly elevated in serum of saline-treated mice as compared to the controls and a significant attenuation of these cytokines were observed in IL233 treated group, almost to the levels of control mice. Minor, but statistically significant elevations in IL-4 were seen for the control out of the three groups. Only saline-treated aGVHD mice exhibited an increase in IL-10 levels however this was not significant when compared to uninjected F1 mice.

Figure 3. IL233-mediated protection involves modification of cytokine milieu.

A) Serum samples were obtained from control, saline and IL233 pre-treated mice at day 14 and levels of cytokines IFNγ, TNFα, IL-4 and IL-10 were quantified using ELISA MAX™ kits. B) Pooled peripheral lymph nodes were obtained at day 14 (terminal endpoint) and were stimulated with PMA/Ionomycin for 5 hours in presence of monensin and expression of IFNγ and TNFα in CD4⁺ and CD8⁺ T cells was analyzed by flow cytometry. Data are shown as mean ± S.E.M and are representative of 6–10 mice per group. * p < 0.01, ^** p < 0.001, ^*** p < 0.0001 as calculated by multiple comparison using one-way ANOVA, Kruskal-Wallis test and Mann-Whitney U test.

Correspondingly, in lymph nodes (Figure 3B), CD4 T helper cells in saline treated aGVHD mice expressed significantly elevated levels of intracellular production of the proinflammatory IFNγ, which was attenuated in IL233 treated mice. Similarly, higher proportion of CD8 T cells in the saline treated mice expressed IFNγ compared to untreated control mice. IL233 treatment mildly decreased IFNγ levels, although this was statistically significant vs. saline treated aGVHD mice. We did not observe significant differences in intracellular TNFα production by either CD4 or CD8 T cells for any of the aGVHD groups. The status of Th1 and other major chemokines (MIG (CXCL9), IP10 (CXCL10), RANTES, MIP-1b, MCP-1, Eotaxin) were also investigated in serum samples. Although there was a trend for lower levels of these chemokines in the IL233 group, the data did not reach statistical significance between the saline and IL233 groups (data not shown). Taken together, the data confirm that IL233 can attenuate the increase in inflammatory cytokines characteristic of aGVHD in this model.

Our study demonstrates a protective effect for IL233 in aGVHD and presents evidence that this effect is mediated in part by regulatory T cells and attenuation of pro-inflammatory cytokines. Our previous studies have shown an induction of a tolerogenic milieu following IL233 treatment as observed by an increase in the M2 macrophages in type-2 diabetes models [4] or lower expression of co-stimulatory molecules in lupus nephritis models [2]. The IL-33/ST2 signaling axis, in particular has been widely investigated with respect to MDSCs (Myeloid Derived Suppressor Cells) for immune modulation and protection from inflammatory conditions [17–20]. In line with these studies, a modulatory role of IL233, if any, on MDSCs may explain Treg-independent mechanisms triggered by IL233 in aGVHD settings. Most importantly, such effects could provide important clues for successful pre-clinical translation and investigations on the effects on MDSCs are in prospects.

4. Conclusions

Our results demonstrate that the hybrid cytokine IL233 can attenuate acute GVHD parameters in the P→F1 murine model. Pretreatment with IL233 was able to impair donor CD8 CTL elimination of host B cells, reduce the reciprocal HVG response, maintain balance between central and effector memory CD8 T-cells and suppress the pathogenic overproduction of inflammatory cytokines associated with aGVHD. These results support further study of IL233 in conditions characterized by either excessive Th1/CD8 CTL function or IFNγ and TNFα secretion.

Highlights:

• An IL-2 and IL-33 fusion cytokine (IL233) attenuated acute GVHD in the parent to F1 murine model.

• IL233 treatment attenuated donor CD8 CTL to host B cells as well as host Vs graft response.

• IL233 restored memory T cell balance and suppressed GVHD-associated proinflammatory cytokines.