The atypical chemokine receptor-2 does not alter corneal graft survival but regulates early stage of corneal graft-induced lymphangiogenesis

Corneal allograft rejection is predominantly mediated through the indirect pathway of allorecognition whereby newly recruited host antigen presenting cells (APC) process and present corneal alloantigens to naïve host T cells. The activation of an allospecific Th1 response then promotes the rejection of corneal allograft [reviewed in [1]]. Unlike allografts, syngeneic corneal grafts performed in naïve hosts are accepted indefinitely [2]. However, despite different outcomes, both corneal syngeneic and allogeneic grafts induce corneal hem- and lymph-angiogenesis which are considered to significantly affect the fate of the graft or the success of a second graft [3, 4].

The atypical chemokine receptor-2 (ACKR2, formerly known as D6) is a chemokine decoy receptor expressed mainly on afferent lymphatic endothelial cells (LEC) and innate-like B cells as well as some other leukocyte subsets such as dendritic cells (DC), macrophages and neutrophils [5] and has also recently been identified on a stromal fibroblastic population in the murine mammary gland [6]. Chemokine decoy receptors, such as ACKR2, have similar structures to conventional chemokine receptors, but they behave anomalously by binding chemokine ligands which fail to initiate G-protein-dependent signalling. Instead, engaged ligands are degraded after internalisation, and, in the case of ACKR2, the receptor is recycled back to the cell surface [5, 7]. ACKR2 recognises most pro-inflammatory CC chemokines with different affinity, but does not recognise homeostatic chemokines [5]. In this way, ACKR2 is regarded as a chemokine scavenger that regulates pro-inflammatory CC chemokine levels which in turn modulate immune responses.

Several studies have shown that ACKR2 is involved in the efficient resolution of inflammation. The absence of ACKR2 leads to more severe inflammatory disease associated with increased chemokine and leukocyte infiltrations [5, 8‐10]. Furthermore, in a previous study, deletion of ACKR2 was reported to be associated with significantly increased rejection of corneal syngeneic grafts with nearly 60% of ACKR2^−/− mice rejecting their syngrafts at 1 week post-surgery [11]. This is a surprising result since ACKR2^−/− mice are fully histo-compatible with wild-type littermate mice and are in effect syngeneic grafts. However, the effect was ascribed to an elevated innate immune response in ACKR2^−/− mice which, under sterile conditions of transplant surgery, implies involvement of damage-associated molecular patterns (DAMPS) [12]. Specifically, it was suggested that ACKR2 expressed by DC plays a role in modifying DC behaviour by promoting maturation and allosensitisation [11]. However, despite showing an impaired allospecific T cell response in ACKR2^−/− mice, allograft survival was not different between WT and ACKR2^−/− mice [11], suggesting that any such effect only affected syngeneic grafts.

An alternative role for ACKR2 function in the innate immune system has recently been suggested. ACKR2 regulates the embryonic development of lymphatic vessels, and that deletion of ACKR2 eventually leads to increased lymphatic vessel density in adult mice in a range of tissues including the skin, diaphragm, and lymph nodes [13]. This phenotype was associated with increased proximity of pro-lymphangiogenic macrophages to LEC in ACKR2^−/− mice [13]. However, it is not clear whether ACKR2 is involved in inflammation-associated lymphangiogenesis in adult mice.

In view of the recognised importance of both chemokines [14] and lymphangiogenesis [4] in corneal graft rejection, we have re-investigated the role of ACKR2 in mouse corneal graft survival using ACKR2-deficient mice. In addition, in view of the proposed role of macrophages and DC in this process [11, 13], we have also evaluated corneal neovascularisation in ACKR2:F4/80 double gene deficient mice. F4/80 is expressed predominantly on resident tissue, suppressive macrophages and a subset of DC [15, 16]. We compared syngeneic, fully MHC incompatible allogeneic and single antigen disparity (HY) graft survival in WT and ACKR2^−/− mice and evaluated blood and lymphatic corneal vessel growth in all groups of mice with syngeneic or allogeneic corneal grafts. We observed no difference in corneal graft survival between WT and ACKR2^−/− in all settings tested while there was only a transient (first 3 days post-graft, [pg]) pro-lymphangiogenic effect on corneal vessels in ACKR2^−/− mice. We are therefore unable to confirm a significant role for ACKR2 in the regulation of innate immune responses in the setting of murine corneal graft or in corneal syn- or allo-graft rejection.

The role of ACKR2 in various inflammatory models is considered to be a general one of limiting excessive inflammation and promoting inflammatory resolution by “scavenging” inflammatory chemokines [5, 8‐10]. Few studies have addressed the role of ACKR2 in allo-immunity [11, 26, 27]. Deletion of ACKR2 had a protective effect in graft-versus-host disease attributed to increased numbers and enhanced immunosuppressive activity of Ly6C^high monocytes [26] while a pro-inflammatory effect was observed in syngeneic but not allogeneic corneal grafts [11]. It was further proposed in this study that ACKR2 promoted DC maturation and T cell activation [11]. Thus, ACKR2 appears to promote innate immunity and in its absence the tempo of corneal allograft rejection might be expected to be reduced, while syngeneic grafts performed under sterile conditions should not be rejected. Counter-intuitively, syngeneic corneal grafts, which are universally accepted in uninfected mice [2, 18, 20] were reported to be rejected in ACKR2^−/− mice while allogeneic corneal graft showed no difference in rejection rates between in WT and ACKR2^−/− mice [11].

In view of the importance of chemokines to the corneal graft rejection process, not least because of possible translational and therapeutic implications, we considered it important to revisit the role of ACKR2 in corneal graft rejection. Using three different donor and recipient combinations (syngeneic, allogeneic and HY-antigen disparity), our data report no difference in corneal graft survival between WT and ACKR2^−/− mice (Figs. 1 and 2). Importantly, in the previous study, C57BL/6 WT donor corneas were used for both WT and ACKR2^−/− mice in syngeneic corneal grafts thus introducing an antigen disparity which may have affected their results [11]. However, the graft rejection rate did not follow the pattern of a single antigen disparity (see Fig. 2 herein) but was rather that of primary endothelial cell failure [18]. Differences in technique may explain the difference between the results of the present study and those of Hajrasouliha et al. [11]. No information is provided on the use of littermate controls in the previous study [11]. In addition, corneal grafts performed using interrupted sutures, as used in the previous study [11], cause significantly greater trauma and subsequent corneal opacity at week 1 pg compared to a continuous suture technique used here [18] and lead to greater inflammation and subsequent stress especially to the corneal endothelial cells [18]. Such technical issues will have a significant impact on innate immune activation.

Although we observed no difference in corneal graft rejection rates linked to ACKR2, the possibility that the process of lymphangiogenesis was altered could not be excluded, particularly since such changes had been observed in embryonic skin tissues in AKCR2^−/− mice [13]. Overall, the level of peri-corneal limbal lymphatic vessels in resting ACKR2^−/− mice was not different from naïve WT mice (Fig. 3a). This finding was not unexpected as the naïve corneal tissues are avascular with a narrow circumferential ring of limbal blood and lymphatic vessels. Whereas shown in the previous study, deletion of ACKR2 altered the density of lymphatic vessel network [13]. However, in agreement with a previous study in F4/80^−/− mice [24], we observed that the levels of blood and lymphatic vessels at the corneal limbus in naïve F4/80^−/−ACKR2^−/− mice were significantly reduced compared to WT and ACKR2^−/− mice (Fig. 3a). This suggests that F4/80⁺ macrophages are likely required for the development of normal corneal limbal vessels. Further support for this concept has been shown in the embryonic skin of WT and ACKR2^−/− mice where two distinct populations of macrophages were identified namely CD11b^hiF4/80^loLyve-1⁻ and CD11b^loF4/80^hiLyve-1⁺ with the latter population expressing higher pro-angiogenic transcripts [13]. Moreover, previously reported experiments of corneal suture-induced lymphangiogenesis also revealed significant suppression of lymphangiogenesis in F4/80^−/− mice compared to WT mice implicating an important role of F4/80⁺ macrophages in corneal lymphangiogenesis [24]. In contrast, we found that after corneal syngeneic graft, F4/80^−/−ACKR2^−/− mice had increased numbers of lymphatic sprouts and increased infiltration of single corneal Lyve-1⁺ cells compared to WT mice (Fig. 3c). Thus, our data suggest that the absence of AKCR2 promoted the recruitment of Lyve-1⁺F4/80⁻ cells which may be capable of promoting the initial lymphatic sprouting and rescued the phenotype of impaired lymphangiogenesis in F4/80^−/− mice as reported before [24]. The process of postnatal lymphangiogenesis is known to involve the proliferation and differentiation of LEC after activation by pro-lymphangiogenic stimuli (e.g., vascular endothelial growth factors) and resulting in the sprouting of existing lymphatics [28]. However, increasing evidence supports the notion that bone marrow-derived lymphatic endothelial progenitor cells might differentiate into LEC and contribute to postnatal de novo lymphovasculogenesis in certain tissues during inflammatory response [23, 29‐31]. Moreover, these cells may also play a paracrine role in promoting lymphangiogenesis by producing pro-lymphangiogenic cytokines [30]. The effect of ACKR2 and F4/80 was transient and restricted to the early stages of injury since no differences were observed at 7d pg in syngeneic grafts (Fig. 3c). It has been previously shown that syngeneic and allogeneic corneal grafts displayed similarly increased levels of chemokines as well as inflammatory cytokines at early time points (within 1 week pg) [32]. Thereafter, chemokine and cytokine levels remained high in the allogeneic group only, whereas in the syngeneic group, these levels declined significantly [32]. Furthermore, our findings coincide with a recent report that deletion of ACKR2 led to accelerated development of mammary gland branching associated with macrophage recruitment [6]. It was further shown that ACKR2 is differentially expressed in the mammary gland during different biological stages with maximum expression correlating with increased branching and macrophage recruitment in ACKR2^−/− mice [6]. In addition, in vitro experiments demonstrated up-regulation of ACKR2 expression upon exposure to IFN-γ and IL-6 [22]. Therefore, ACKR2 in the current model is likely to function by regulating levels of inflammatory chemokines which in turn affects the recruitment of pro-lymphangiogenic Lyve-1⁺ cells during early time points (3d pg) after syngeneic grafts rather than allogeneic grafts. However at later time points where chemokine levels fall significantly in WT mice, ACKR2 may be less effective. Thus, together with our data, suggests that while ACKR2 may exert variable function during different stages of inflammation, we show here that ACKR2 plays a specific regulatory role of early inflammation-associated lymphangiogenesis in adult mouse.

Interestingly, this effect of ACKR2 on lymphangiogenesis was observed in early stages of inflammation of syngeneic grafts but not allogeneic grafts (Fig. 3e). Since the alloimmune response is more prominent and long lasting compared to the immune response to syngeneic grafts (which is, in effect, similar to an autologous corneal wounding response), we suggest that the strength of alloimmune response may mask a more subtle effect on simple wound healing responses. Alternatively, there may be an early suppression of lymph vessel sprouting mediated by the adaptive alloimmune response. This may merit further investigations.

In summary, we report that in contrast to a previous study [11], ACKR2 does not have a role to play in regulating corneal allograft responses. There is a minor effect mediated by ACKR2 during the early stages of lymphangiogenesis in corneal wound healing type responses, but this effect does not change final outcome of the transplanted corneal grafts.