Hybrid resistance to parental bone marrow grafts in nonlethally irradiated mice

Resistance to parental bone marrow (BM) grafts in F1 hybrid recipients is due to natural killer (NK) cell–mediated rejection triggered through “missing self” recognition. “Hybrid resistance” has usually been investigated in lethally irradiated F1 recipients in conjunction with pharmacological activation of NK cells. Here, we investigated BM‐directed NK‐cell alloreactivity in settings of reduced conditioning. Nonlethally irradiated (1‐3 Gy) or nonirradiated F1 (C57BL6 × BALB/c) recipient mice received titrated doses (5‐20 x 106) of unseparated parental BALB/c BM without pharmacological NK cell activation. BM successfully engrafted in all mice and multilineage donor chimerism persisted long‐term (24 weeks), even in the absence of irradiation. Chimerism was associated with the rearrangement of the NK‐cell receptor repertoire suggestive of reduced reactivity to BALB/c. Chimerism levels were lower after transplantation with parental BALB/c than with syngeneic F1 BM, indicating partial NK‐mediated rejection of parental BM. Activation of NK cells with polyinosinic–polycytidylic acid sodium salt poly(I:C), reduced parental chimerism in nonirradiated BM recipients but did not prevent hematopoietic stem cell engraftment. In contrast, equal numbers of parental lymph node cells were completely rejected. Hence, hybrid resistance leads to incomplete rejection of parental BM under reduced conditioning settings.


| INTRODUC TI ON
Natural killer (NK) cells are large granular lymphocytes that serve as first-line defense against pathogens and neoplastic cells. 1 The expression of germline-encoded receptors endows them to discriminate between healthy, infected, and malignant cells. It is also widely acknowledged that NK cells play a major role in the rejection of allogeneic bone marrow (BM), although they are not able to fully reject solid allografts. 2,3 The importance of NK cells in the rejection of allogeneic BM cells has been elegantly demonstrated when parental BM was transplanted into the first generation (F1) of offspring recipients ("hybrids"). 4 Under these circumstances Resistance to parental bone marrow (BM) grafts in F1 hybrid recipients is due to natural killer (NK) cell-mediated rejection triggered through "missing self" recognition. "Hybrid resistance" has usually been investigated in lethally irradiated F1 recipients in conjunction with pharmacological activation of NK cells. Here, we investigated BM-directed NK-cell alloreactivity in settings of reduced conditioning. Nonlethally Chimerism was associated with the rearrangement of the NK-cell receptor repertoire suggestive of reduced reactivity to BALB/c. Chimerism levels were lower after transplantation with parental BALB/c than with syngeneic F1 BM, indicating partial NK-mediated rejection of parental BM. Activation of NK cells with polyinosinicpolycytidylic acid sodium salt poly(I:C), reduced parental chimerism in nonirradiated BM recipients but did not prevent hematopoietic stem cell engraftment. In contrast, equal numbers of parental lymph node cells were completely rejected. Hence, hybrid resistance leads to incomplete rejection of parental BM under reduced conditioning settings.

K E Y W O R D S
animal models: murine, basic (laboratory) research/science, bone marrow/hematopoietic stem cell transplantation, immunobiology, innate immunity, natural killer (NK) cells/NK receptors, rejection, tolerance: chimerism in which T cells exhibit no anti-donor alloreactivity, the lack of recipient MHC molecules on donor cells triggers "missing-self recognition" by recipient NK cells and leads to rejection of the parental BM ("hybrid resistance"). [5][6][7] This model has ever since been a valuable tool for elucidating the role of NK cells in allogeneic BM transplantation but also has its limitations. In the classical hybrid resistance model, recipients are lethally irradiated and NK cells are actively preactivated with polyinosinic-polycytidylic acid sodium salt (poly(I:C)). 4,8 This model was developed in the early 1970s 6 when myeloablative doses of irradiation were routinely used for BM transplantation. In recent decades, reduced conditioning regimens based on nonlethal irradiation have gained importance in the clinical setting of BM transplantation, but the effect of NK-mediated hybrid resistance remains unclear under these conditions.
To address this issue, we investigated hybrid resistance under reduced conditioning settings.

| Skin transplantation
Full-thickness tail skin was grafted 4-6 weeks after BM transplantation and visually inspected thereafter at short intervals. Grafts were considered to be rejected when less than 10% remained viable, as described earlier. 10

| Flow cytometry
The presence of donor cells was assessed at regular intervals by staining CD45.1 and CD45.2 on blood leukocytes. Donor chimerism was assessed as percentage of CD45.1 − CD45.2 + cells among

| Statistical analysis
Data were statistically analyzed with GraphPad Prism 5.0 (Graph Pad Inc., La Jolla, CA). A 2-sided Student's t test with equal variances was used to compare chimerism levels. Total chimerism levels were compared between groups by using analysis of variance (ANOVA).

| RE SULTS
To track BM engraftment and chimerism for an extended period, we NK cells can simultaneously express inhibitory receptors (Ly49A, Ly49G2) that bind the very same MHC molecule. 13 Those NK cells that express the activating receptor Ly49D without expressing any of the inhibitory receptors Ly49A or Ly49G2 are potentially donorreactive 9,14 ( Figure 1F). Transplantation of BALB/c BM into nonirradiated F1 mice significantly reduced the amount of Ly49D + Ly49A/ G2 − NK cells ( Figure 1G, H). The rearrangement of the NK cell receptor repertoire evolved over the first 4 weeks posttransplant and remained stable thereafter ( Figure 1I). NK cell adaption did not occur if allogeneic skin or syngeneic BM was transplanted ( Figure 1J). The degree of NK cell receptor rearrangement was independent of the dose of irradiation and the ensuing levels of chimerism ( Figure 1K).
Transplantation of parental BM altered the appearance of inhibitory receptors but had no effect on the expression of the activating receptor Ly49D ( Figure 1L). Thus, in recipients receiving no or nonlethal irradiation, NK cells did not abrogate engraftment of parental BM, but rather adapted through the rearrangement of their receptor repertoire. 15 Next, we decreased the numbers of BALB/c BM cells transplanted into nonirradiated F1 mice. Donor leukocyte chimerism was detectable long-term with all applied BM doses, even with the lowest dose of 5 × 10 6 cells (Figure 2A). Multi-lineage chimerism, however, developed only with a BM dose of 10 × 10 6 or higher, as no CD3 or CD19 chimerism was detectable with 5 × 10 6 cells ( Figure 2B, C).
Overall, the dose of transplanted BM cells and the level of leukocyte chimerism showed a linear correlation ( Figure 2D). To determine whether partial rejection of parental BM cells occurs in successful chimeras, we compared chimerism levels between recipients of pa-

| D ISCUSS I ON
The results presented herein reveal that under reduced conditioning settings, parental BM is only partially rejected by NK cells in F1 recipients. Multi-lineage chimerism ensues even in nonirradiated F1 recipients transplanted with moderate BM doses. Complete rejection of parental BM is not triggered, even when poly(I:C) is given.
Irradiation promotes engraftment by creating space in the BM niche, 18 but also leads to the release of proinflammatory cytokines 19,20 and other danger signals that are expected to enhance NK alloreactivity. 1,21 Therefore it is tempting to speculate that NK cell alloreactivity was mitigated under reduced intensity conditioning. Pharmacological NK stimulation with poly(I:C) led to complete rejection of parental lymphocytes but not parental BM, allowing stem cell engraftment in this setting. The reasons why NK cells preferentially target lymphohematopoietic cells remain unclear but likely reflect the selective expression of distinct receptors. 22 It has also been suggested that NK cells exhibit no direct cytotoxicity against stem cells 17,23 and that they reside within immune privileged sites that prevent them from undesired immune attack. 24 So far the proliferation of recipient splenocytes shortly after BM transplantation served as surrogate marker for BM engraftment in hybrid resistance models using lethally irradiated mice. 4 This end point, however, does not allow drawing robust conclusions regarding long-term chimerism. 25 The model presented herein provides a new oppurtunity to follow parental donor cells in F1 recipients by flow cytometry and is thus particularly suited for the investigation of NK-mediated hybrid resistance under distinct reduced conditioning settings. Our results demonstrate that NK-mediated rejection of parental BM is diminished and remains incomplete in nonlethally irradiated recipients. Long-term multi-lineage chimerism was observed even in nonirradiated recipients. This finding also indicates that "space" does not necessarily have to be created in the recipient through myelosuppression for hematopoietic stem cells to engraft. This has already been suggested previously in models of high-dose BM administration, 26,27 and recently it has been reported that ample free sinusoidal perivascular niches exist where exogenous stem cells can engraft. 16 Even if allogeneic stem cells have sufficient space to engraft, one would expect NK cells to resist their engraftment unless very large BM doses are infused. 15 Unexpectedly, NK adaptation occurred at moderate BM doses, reminiscent of the NK adaptation seen with the chronic exposure of viruses that is associated with decreased expression of the activating receptor Ly49H. 28 However, we did not observe alterations in the expression of the donor-specific activating receptor Ly49D in established mixed chimeras. It rather seemed that Ly49D + NK cells would obtain the expression of the inhibitory receptors Ly49A and/or Ly49G2. This adaptation extended over a period of 4 weeks, which approximately corresponds to the time of NK cell maturation in the BM. 29 The altered expression of Ly49 inhibitory receptors in MHC class I deficient mice and in fully allogeneic BM chimeras further supports this assumption. 14,30 Our data from the murine hybrid resistance setting suggest that NK-mediated BM rejection is less potent in reduced conditioning settings than in lethal irradiation regimens, allowing stem cell engraftment with moderate BM doses even in nonirradiated recipients.