Female C57BL/6 (B6, recipient, H-2^b), BALB/c (donor, H-2^d), and C3H/N (third party, H-2^k) mice were purchased from Charles River Laboratories (Sulzfeld, Germany), housed under specific pathogen-free conditions, and used for experiments between 6 and 12 weeks of age. Congenic B6.SJL-Ptprca Pepcb/BoyJ (CD45.1⁺) and B10.D2 mice [donor, MHC mismatched, minor antigen (mHAg) matched to B6] were purchased from Jackson Laboratory (ME, USA) and bred in our own facility. Skin and heart transplantations were performed under controlled anesthesia with an intraperitoneally (i.p.) injected mixture of xylazin (5 mg/kg) and ketamine (100 mg/kg). All experiments were approved by the local review board of the Medical University of Vienna and the Austrian Federal Ministry of Science, Research and Economy and were performed in accordance with national and international guidelines of laboratory animal care (permission number GZ: BMWFW-66.009/0377-WF/V/3b/2016 and GZ: BMWFW-66.009/0303-WF/V/3b/2016).

Groups of age-matched C57BL/6 recipients received 20 × 10⁶ unseparated BM cells or splenocytes [as donor-specific transfusion (DST) control] from BALB/c donors by injection into the tail vein together with CB consisting of anti-CD154 mAb (1 mg, d0, anti-CD40L, MR1, BioXcell), hCTLA4-Ig (0.5 mg, d2, abatacept, Bristol Myers Squibb), and a short-course of rapamycin (0.1 mg, d-1, d0, and d2, LC Laboratories) (6). Indicated recipients were additionally treated with an anti-IL-6 mAb (1 mg on days −1, 1, and 3 and then 0.1 mg every other day until day 13, MP5-20F3, BioXcell) (27, 28). Where indicated, groups of mice received in addition (0.5 mg on days −1, 1, and 3 and 0.1 mg every other day until day 13) anti-IL-6R (15A7) or anti-IL-17A (BZN035, Novartis Pharma AG) (42, 43).

Full thickness tail skin from BALB/c, B10.D2, or C3H/N mice was grafted to recipient flanks (lateral thorax wall) 5 weeks after BMT and visually inspected thereafter at short intervals. Grafts were considered to be rejected when less than 10% of the graft remained viable (6). Cervical heterotopic heart transplantation was performed 5 weeks after BMT, as described previously (44). Briefly, the recipient’s right external jugular vein and common carotid artery were everted over a cuff. The donor heart was harvested and flushed in a retrograde fashion with 4 mL HTK solution (Custodiol, Koehler Chemie, Alsbach-Haenlien, Germany) through the aortic arch. The pulmonary artery was connected with the external jugular vein and the aortic trunk with the common carotid artery. Heart allograft survival was determined by visual inspection and palpation at least twice weekly during long-term follow-up. End of graft survival was defined as complete cessation of heartbeat and was confirmed by histological analysis.

Skin and heart grafts were harvested at the end of the experiment and fixed in 7.5% formalin. Paraffin blocks were sectioned and stained with hematoxylin and eosin (H&E) according to standard protocols. Histological slides were then scanned by Aperio ScanScope scanner (Aperio Technologies, Inc., Vista, CA, USA) and analyzed with ImageScope software. Additionally, grading of skin grafts was performed according to Banff 2007 working classification of skin-containing composite tissue allograft pathology and hearts were scored according to the International Society for Heart and Lung Transplantation (ISHLT) 2005 guidelines for cellular rejection by a pathologist blinded to the experimental background of samples.

Regarding MLR 5 × 10⁵ congenic C57BL/6 CD45.1⁺ responder cells (unseparated splenocytes) were cultured for 5 days in RPMI1640 supplemented with 10% FCS (Linaris), PenStrep (100 U Penicillin, 100 µg Streptomycin/ml; Sigma), 10 mM Hepes (MP Biomedicals), 1 mM Sodium Pyruvat (Sigma), and 1× non-essential amino acids (Sigma). Responder cells were stimulated with equal numbers of stimulator cells (unseparated splenocytes) from BALB/c (allogeneic) and C57BL/6 (self) or with Medium. Proliferation was measured by staining Ki-67 in responding (CD45.1⁺) CD4⁺ and CD8⁺ T cells after 5 days of culture. For the in vitro suppression assay, a 96-well plate was pre-coated overnight with 1 µg/ml anti-CD3 (145-2C11, BioXcell) diluted in PBS. On the next day, 4 × 10⁵ naïve responder CD45.1⁺ splenocytes (VPD 450 pre-labeled) were cocultured with indicated numbers of Tregs (for a responder:Treg ratio of 1:0, 1:1, 2:1, 4:1, 8:1, and 16:1) isolated on day 14 from BMT recipients (receiving CB and rapamycin) either with or without anti-IL-6. Tregs were purified by magnetic bead separation using the MACS CD4⁺CD25⁺ regulatory T-cell Isolation Kit (Miltenyi Biotec). Purity of separated population was >90%. After 4 days of incubation responder cell proliferation was assessed by measuring the dilution of VPD450. The percentage of suppression was calculated using the formula: [(% proliferation Tresp. cells alone − % of Tresp. cells treated with Treg)/% of Tresp. cells alone] × 100.

Three-color flow cytometric analysis of multi-lineage chimerism was performed as described previously (45, 46). Briefly, chimerism was calculated as the net percentage of donor MHC class I⁺ [H2-D^d-BIO (34-2-12)] in different leukocyte lineages {myeloid cells [Mac-1-FITC (M1/70)], B cells [CD19-PE (6D5)], CD4⁺ cells [CD4-APC-Cy7 (RM4-5)], and CD8⁺ cells [CD8-PE-Cy7 (53-6.7)]}. CD44-BIO (IM7), CD62L-PE (MEL-14), CD11c-BIO (N418), CD80-PERCP-Cy5 (16-10A1), CD86-FITC (GL-1), MHCII-PE-Cy7 (M5/114.15.2), CD40-PE (3/23), CD3-PerCP-Cy5 (17A2), SAV-APC, CD45.1-BIO (A20), CD45.2-BIO (104), and IgG-PE (Poly4053) were purchased from BioLegend (SanDiego, CA, USA). For intracellular staining, cells were permeabilized with the Foxp3/Transcription Factor Staining Buffer Set from eBioscience according to the manufacturer’s specification. Anti-mouse/rat Foxp3 APC (FJK-16s) and anti-mouse/rat Ki-67 PE-Cy7 (SolA15) were obtained from eBioscience. Flow cytometric analysis was performed with a BD FACS Canto II, and data were analyzed using FlowJo (10.0.8) software.

5 × 10⁵ naïve recipient or donor thymocytes were incubated with 15 µl serum of indicated mice for 30 min at 37°C after heat inactivation of complement as described earlier (6). Briefly, binding of donor-reactive IgG to thymocytes was assessed by staining with anti-IgG (PE) and was measured by flow cytometry.

The statistical analyses were performed with SPSS 23.0 and GraphPad Prism 5.0 (GraphPad Software, Inc., La Jolla, CA, USA). Differences between chimerism levels were compared using ANOVA. Error Bars represent SD and statistics were performed with a two-sided Student’s t-test with equal variances. Skin and heart allograft survival was calculated according to the Kaplan–Meier product limit method and compared between groups using the log-rank test. Cardiac rejection scores were compared by using Fisher’s exact test. A p-value < 0.05 was considered statistically significant.

To investigate whether IL-6 blockade can replace Treg cell therapy in an irradiation-free BMT setting, C57BL/6 mice received fully mismatched BALB/c BM, CB (anti-CD40L mAb and CTLA4-Ig), rapamycin and in addition anti-IL-6 (instead of adoptive Treg transfer). All mice receiving anti-IL-6 developed early multi-lineage chimerism (within the myeloid and B cell lineages) that persisted for up to 12 weeks (18/18 chimeric at week 2, 6/18 mice remained chimeric at week 12, pooled results from three independent experiments). Noticeably, T cell chimerism appeared in most anti-IL-6 treated mice between weeks 6 and 9 and lasted until week 12 (Figures 1A,B), whereas no T cell chimerism was detectable in mice without anti-IL-6 (week 9: 14/18 with anti-IL-6 vs 0/10 without, p = 0.017). Among all measured leukocyte lineages, the myeloid population showed the highest chimerism levels over the whole period, peaking at week 2. In contrast, almost all control mice without anti-IL-6 (and without Treg transfer) lost myeloid (and B cell, data not shown) chimerism by week 3, or shortly thereafter (week 3: 1/10 chimeric vs 18/18 chimeric with anti-IL-6, p = 0.021) (Figure 1C), consistent with our previous experience (6).

To determine whether anti-IL-6 leads to donor-specific tolerance, donor skin was transplanted ~5 week’s post-BMT (i.e., 22 days after the last dose of anti-IL-6). Donor skin graft survival was significantly prolonged in anti-IL-6-treated mice [median survival time (MST) 58 days, n = 18] compared to control mice without anti-IL-6 (MST 24 days, n = 10, p = 0.0001). Third-party skin grafts were rapidly rejected in all groups (MST 11 and 10.5 days, respectively). To test whether the observed prolongation of BALB/c donor skin graft survival after BMT with anti-IL-6 depends on (transient) BM engraftment, a group of mice was treated with the same anti-IL-6 regimen but received donor splenocytes instead of BM (DST α-IL-6, n = 4). DST recipients promptly rejected donor skin (MST = 12.5 days), suggesting that BM engraftment is essential for graft prolongation and that DST does not suffice (Figure 1D). No DSA were detectable in sera of anti-IL-6 treated recipients 100 days post skin grafting, while more than half of mice without anti-IL-6 presented detectable levels of DSA (0/8 vs 5/8, p = 0.019), indicating that donor-specific humoral immunity is prevented in anti-IL-6 treated mice (Figures 1E,F).

Minor histocompatibility antigens (mHAg) are a barrier to tolerance induction through chimerism (47, 48). To investigate whether mHAg disparities are responsible for the late rejection of donor skin grafts in anti-IL-6-treated recipients, B10.D2 skin was grafted instead of BALB/c skin. Survival of B10.D2 grafts was significantly prolonged compared to BALB/c grafts (MST 86 vs 58 days, p = 0.0014) but all grafts were eventually rejected (Figure 1G), suggesting that mHAg disparities are not solely responsible for late rejection.

These results show that anti-IL-6 prolongs donor skin graft survival due to enhanced allogeneic BM engraftment.

Skin grafts are generally considered to be the most stringent test for tolerance, while vascularized heart grafts are better suited to assess chronic rejection. Therefore, additional groups were transplanted with donor hearts 5 weeks post-BMT. Remarkably, in anti-IL-6-treated BMT recipients all donor hearts remained viable with a palpable heart beat for the length of the follow-up (n = 4, 100 days), whereas 3/4 recipients from the group without anti-IL-6 rejected their grafts (MST 88.5, p = 0.04) (Figure 2A). Histopathologic examination revealed lower rejection scores in mice treated with anti-IL-6 [median ISHLT score 2 (anti-IL-6) vs 4 (w/o anti-IL-6)] (Figure 2C). Heart grafts from mice without anti-IL-6 showed severe signs of cellular rejection, including extensive lymphocytic infiltrations, myocardial fibrosis, and necrosis, while three of four grafts from anti-IL-6 treatment demonstrated only mild to moderate signs of rejection, including lower rates of pericardial inflammation and intima arteritis (Figure 2B).

Hence, these data demonstrate that anti-IL-6 treatment leads to long-term acceptance of donor heart grafts in an irradiation-free BMT setting.

Regulatory T cells modulate immune responsiveness to alloantigens, thereby modulating graft survival (49). To test whether the chimerism- and tolerance-promoting effects of anti-IL-6 administration are associated with in vivo boosting of endogenous Tregs, we measured the frequency, number, and proliferation of Foxp3⁺CD4⁺ T cells. Anti-IL-6 was associated with significantly increased frequencies of Foxp3 Tregs in blood (on days 14 and 21 post-BMT) (Figures 3A,B) and spleen (not shown), with an approximately 30% increase in absolute Treg numbers in spleen (Figure 3C). Moreover, Treg proliferation was elevated, as measured by the proliferation marker Ki-67 (Figures 3D,E). As it has been reported that CD4⁺CD25⁺ Tregs isolated from IL-6-deficient mice are more potent than those from WT mice (27), an in vitro suppression assay with polyclonally activated CD45.1⁺ splenocytes as responders and descending numbers of CD45.2⁺CD4⁺CD25⁺ Tregs isolated from BMT recipients treated either with or without anti-IL-6 was performed. Tregs from both groups suppressed the proliferation of responder T cells to a comparable degree in a dose-dependent manner (Figure 3F).

These data provide evidence that anti-IL-6 therapy leads to a quantitative expansion of endogenous Tregs, but does not qualitatively alter their suppressive potency.

Dendritic cells play a central role in alloimmunity through the presentation of alloantigens to naïve and memory T cells. IL-6 influences the maturation and activation of DCs (23). Therefore, recipient spleens of anti-IL-6-treated and untreated recipients were isolated on day 7 and 14 post-BMT and the DC population (CD45⁺CD11c⁺CD3⁻) was analyzed by flow cytometry. The number of DCs was significantly reduced in anti-IL-6-treated mice compared to the untreated group (~20% less at day 14, p = 0.018) (Figure 4A). Moreover, anti-IL-6-treated mice showed a significantly lower expression of CD80 and CD86 (anti-IL-6 CD86: 12%, CD80: 15% vs without anti-IL-6 CD86: 17% and CD80: 21%, p ≤ 0.05), suggesting reduced DC activation (MHCII and CD40 were similarly expressed in both groups) (Figures 4B,C).

Interleukin-6 is also important for the expansion and survival of effector T cells and the persistence of immunological memory (22). Thus, we followed the development of memory T cell subsets over time. In mice treated with anti-IL-6 the frequency of naïve CD8⁺ T cells (CD62L^high/CD44^low) remained constant at around 60% (similar to pre BMT levels), whereas it declined in mice not receiving anti-IL-6 (Figures 4D,F). In parallel, the frequency of CD8⁺ effector memory T cells (T_EM, CD62L^low/CD44^high) remained relatively stable in the anti-IL-6 group and was significantly lower compared to the group without anti-IL-6 with the most pronounced difference at week 3 (~30% reduction, p ≤ 0.001) (Figures 4E,F). In contrast, generation of memory CD4⁺ T cells was not affected (data not shown). To investigate whether the tempered CD8⁺ T_EM development affects the “second-set” allograft response, a second donor skin graft was transplanted 3 weeks after the rejection of the first donor graft. Both groups rejected second donor skin grafts with similar kinetics as first grafts were rejected in untreated mice (Figure 4G). 2 weeks after second graft rejection low but detectable DSA levels appeared in anti-IL-6-treated mice (3/5), which were absent after rejection of the first graft (0/8, p = 0.039) (Figure 4H). In addition, an in vitro MLR with splenocytes, harvested from BM recipients after rejection of the first skin graft, revealed no significant difference regarding proliferation of both CD4⁺ and CD8⁺ T cells whether recipients were treated with anti-IL-6 or not, which is consistent with the in vivo results of skin graft rejection (Figure 4I).

Therefore, we conclude that anti-IL-6 reduces the expansion and activation of DCs and impairs CD8⁺ T_EM generation during BMT. However, this is not sufficient to delay rejection of a second skin graft.

Anti-IL-6 prolongs, but does not extend skin graft survival indefinitely in our model. Previously it has been demonstrated that IL-6 KO and IL-6R KO mice display phenotypic differences which might be explained by the fact that soluble IL-6R shares 60% homology with the IL-12p40 subunit which results in low-affinity interaction with IL-12 and IL-23 (50, 51). Furthermore, distinct outcomes were seen in a murine contact burn model where combinational treatment of anti-IL-6 and anti-IL-6R showed a significant reduction of the inflammatory response compared to anti-IL-6 or anti-IL-6R alone (52). Besides IL-6, increased serum IL-17 levels could be found in several murine acute and chronic allograft rejection models and it was shown that elevated IL-17 levels promote rejection in part by suppressing Treg expansion (53, 54). Thus, we asked whether the combination with anti-IL-17 or anti-IL-6R provides an additional benefit to anti-IL-6 treatment with regard to inducing long-term skin graft acceptance.

Anti-IL-17 and anti-IL-6R further increased the percentage of blood Tregs on day 21 post-BMT when added to anti-IL-6 treatment (anti-IL-17: p* = 0.039 and anti-IL6R: p = 0.059; n = 6/grp) (Figure 5A). In contrast, anti-IL-17 alone showed no difference compared to anti-IL-6 (data not shown). Myeloid chimerism levels were numerically higher at all measured time points in mice treated with anti-IL-6 and anti-IL-17 compared to mice treated with anti-IL-6 alone (Figure 5B), with the clearest difference at week 6 (Figure 5C). There was, however, only a slight, non-significant prolongation of donor skin graft survival with addition of anti-IL-17 or anti-IL-6R (Figure 5D).

Taken together, these data suggest that the combination of anti-IL-6 and anti-IL-17 further expands endogenous blood Tregs and slightly improves BM engraftment, but is not sufficient to achieve permanent donor skin graft acceptance.