NK cells reject non-self hematopoietic bone marrow (BM) grafts via Ly49 receptor-mediated MHC class I-specific recognition and calibration of receptor expression levels. In this paper we investigated how Ly49+ subset frequencies were regulated dependent on MHC class I expression. The development of donor and host Ly49A+ (recognizes H-2Dd and H-2Dk ligands) and Ly49C/I+ (Ly49CBALB/c recognizes H-2Kb, H-2Kd, and H-2Dd, and Ly49CB6 recognizes only H-2Kb) NK cell frequencies were monitored for 120 days in murine-mixed allogeneic BM chimeras. C57BL/6 (H-2b) BM was transplanted into BALB/c (H-2d) mice and vice versa. Peripheral NK cell populations were examined every 5 days. Chimerism was found to be stable with 80–90% donor NK cells. In contrast to syngeneic controls reexpressing pretransplant patterns, donor and host NK cells revealed new and mainly reduced subset frequencies 55 days after allogeneic transplantation. Recipient NK cells acquired these later than donor NK cells. In H-2d → H-2b chimeras Ly49A+, Ly49C/I+, and Ly49A+/Ly49C/I+ proportions were mainly diminished upon interaction with cognate ligands. Also in H-2b → H-2d chimeras, Ly49A+ and Ly49A+/Ly49C/I+ subsets were reduced, but there was a transient normalization of Ly49C/I+ proportions in the noncognate host. After 120 days all subsets were reduced. Therefore, down-regulation of developing Ly49A+ and Ly49C/I+ chimeric NK cell frequencies by cognate ligands within 7–8 wk after BM transplantation may be important for successful engraftment.

In recent years it has become clear that NK cells are important contributors to the rejection of MHC class I-different bone marrow (BM)3 grafts (1, 2). This includes the rejection of parental marrow by the F1 hybrid generation (3, 4) and the rejection of MHC class I-deficient (β2-microglobulin (β2m) mutant) marrow by normal mice (5, 6, 7, 8) and parental marrow (9) or in vitro target cells by MHC-transgenic mice (10) as well as by mice transgenic for human killing inhibitory receptors (KIRs) (11, 12). NK cells are radioresistant and therefore survive lethal irradiation and kill the non-self graft targets according to the missing self hypothesis (13, 14) and the external/internal calibration model (15). When self-MHC class I is missing, the NK cell is not inhibited via a Ly49 inhibitory receptor and thus activated. Defense of non-self MHC is facilitated by down-regulation of receptor expression levels on the cell surface and calibration against self-MHC during ontogeny. But in addition, there is growing evidence for a more diverse role of NK cells, as they can also foster the engraftment process (16). The mechanisms are still poorly understood. One assumption is that KIRs of the Ly49 family of C-type lectin proteins (17) may be of importance. For example, NK cells expressing Ly49C improve the growth of CFUs of syngeneic cells in vitro (18). This improvement seems somewhat surprising considering the better known killing function via Ly49 inhibitory receptors, which only mediate a protective signal upon encountering self-MHC class I but give way to lysis of the target if not inhibited.

At least nine genes of the mouse Ly49 multigene family have been identified so far (19). The Ly49 molecules are thought to play a decisive role in determining the specificity of NK cells through recognition of class I molecules and subsequent delivery of inhibitory and in two cases, Ly49D and Ly49H (20, 21), activitory signals. The cognate Ags for Ly49A are H-2Dd and Dk. Ly49C/I recognizes the Ags Kb, Kd, and Dd. Ly49C alone of C57BL/6 (B6) mice recognizes H-2Kb whereas Ly49C of BALB/c mice recognizes H-2Kb but also H-2Kd and H-2Dd (22, 23). The ligand for Ly49I is still unknown. Ly49I is not expressed by BALB/c mice. The Ly49 molecules define separate NK cell subsets but are partially coexpressed and form overlapping subsets (24); other NK cells do not express known Ly49 at all. The important question of how NK cells acquire specificity and tolerance during maturation and within a naturally MHC-varying environment is still unanswered. Evidence for an influence by host MHC class I has been put forward (14, 25, 26, 27, 28). It seems inconsistent that, for example, very similar frequencies of cells expressing a certain receptor, even a potentially harmful one, are observed in completely MHC class I-different mice (14). It is postulated that after an initial random expression of Ly49 receptors (29) and generation of a diverse repertoire (30), cells are selected via MHC expression levels of the host (31). Potentially harmful cells are either deleted by clonal selection or adapt by becoming anergic or having an altered specificity (13). It was shown that Ly49A+ NK cells from H-2b mice showed reduced killing of H-2b Con A blasts compared with those from H-2d mice but showed efficient killing of β2m lymphoblasts (13).

The production of fully allogeneic MHC-disparate, for example H-2d and H-2b, chimeric mice poses a challenge to donor and recipient NK cells developing in mixed MHC class I hematopoietic hosts. Changes of expression levels in this context rather than expression frequencies have been described so far (32, 33). Recently the numbers of Ly49A+, Ly49C+, and Ly49G2+ NK cell subsets were also addressed in [BALB/c → B6] chimeras (34) at 6 wk after BM transplantation (BMT).

In this study, we examined the in vivo development of Ly49+ subset frequencies of peripheral NK cells at 5-day intervals to determine alterations over time within the critical phase of engraftment. We also analyzed whether these newly acquired patterns remained stable in the surviving mice. We focused on Ly49A and Ly49C/I receptors, analyzing donor and host NK cells separately in [BALB/c → B6] as well as [B6 → BALB/c] chimeric mice. The Ly49 receptor repertoire alterations found are discussed in the context of the external/internal MHC class I-dependent calibration model recently proposed by Kåse et al. (15). Hence, the present study is an approach toward understanding NK cell tolerance in an allogeneic BMT model and thus might provide a basis for further improvement of the engraftment process.

Normal inbred BALB/c (H-2d) and C57BL/6 (H-2b) mouse strains were used as an in vivo system for allogeneic and syngeneic BMT. BALB/c and B6 mice were bred in the animal facilities of Hannover Medical School (Hannover, Germany), maintained under germ-free conditions, and used at the age of 8–12 wk. An oral antibiotic prophylaxis with sulfamethoxazole and trimethoprim in the drinking water of donor and recipient mice was started 3–4 days before irradiation and continued for 60 days. Animals with four indicators of graft-vs-host disease (GvHD) (alopecia, scleroderma, weight loss, hunched posture, severe diarrhea, bleeding, or inflammation of the eyes or nose) were suspected of presenting GvHD and underwent histological examination.

Currently available Ly49 mAbs (35) such as FITC- or PE-conjugated anti-Ly49A-FITC (A1), anti-Ly49C/I-PE (5E6) as well as anti-NK1.1-FITC and -PE (PK136), anti-DX5-FITC and -PE (DX5, a pan-NK cell marker), anti-H-2Kb (AF6–88.5), and anti-H-2Kd (SF1–1.1) were purchased from PharMingen (Hamburg, Germany). Ly49C/I mAb recognizes both Ly49C and Ly49I (36). A1 does not label the BALB/c allele of Ly49A but labels the B6 allele (21). B6 NK cells were distinguished from BALB/c cells by using anti-H-2Kb as well as anti-NK1.1 reacting with B6 but not BALB/c NK cells. Isotype-matched mouse IgG2a mAbs were used as negative staining controls. To block FcγII/III receptor-mediated unspecific binding, anti-FcγII/III (2.4G2) from purified hybridoma supernatants was used. The Ab-producing hybridoma was obtained from American Type Cell Collection (Manassas, VA). To eliminate T lymphocytes in the graft, anti-Thy-1.2 was applied as hybridoma supernatant (the hybridoma was kindly provided by S. Izui, Geneva, Switzerland) as well as low toxicity rabbit complement, purchased from Cedarlane Laboratories (Hornby, Ontario, Canada).

BM cells were flushed out from femur and tibia of two donor animals per five recipients under aseptic conditions. Stem cells were enriched via a Ficoll-gradient, washed in Dulbecco’s PBS, and depleted of T cells by incubation with anti-Thy-1.2 followed by low toxicity rabbit complement according to the manufacturer Cederlane Laboratories. Cell suspensions were adjusted with Dulbecco’s PBS to ∼5–8 × 106 BM cells/inoculum and kept on ice. The recipient B6 mice were exposed to a lethal dose of 1200 rad of gamma-irradiation from a Cs137 source, given in two doses of 600 rad each, 5–6 h apart. The recipient BALB/c mice were exposed to a lethal dose of 950 rad in two doses of 475 rad. Within the next 2 h, they were inoculated with syngeneic and allogeneic BM cells of sex- and age-matched B6 and BALB/c mice, respectively (designated [donor → recipient]). After i.p. narcotization, the BM cell suspensions were injected into the tail veins. The relatively small number of 5–8 × 106 graft cells was chosen in order not to override the recipient’s reaction (37) and to achieve a stable 80–90/10–20% donor/recipient chimerism.

B6 mice were transplanted with allogeneic BALB/c BM (n = 15; [BALB/c → B6]) or syngeneic B6 BM (n = 10; [B6 → B6]). BALB/c mice were transplanted with allogeneic B6 BM (n = 13; [B6 → BALB/c]) or syngeneic BALB/c BM (n = 10; [BALB/c → BALB/c]). The syngeneic BMTs served as controls. All experiments were performed twice with five (eight in one case) recipients each. The [BALB/c → B6] BMT was performed three times.

After syngeneic [B6 → B6] BMT, 8 of 10 mice (80%) survived long-term (>120 days); one death after 40 days was unrelated to BMT. Five of 10 mice (50%) survived long-term after syngeneic [BALB/c → BALB/c] BMT. After allogeneic [BALB/c → B6] BMT, 4 of 15 mice (27%) survived >120 days; 33% had been alive until day 40. Six of 13 (46%) BALB/c mice survived long-term after allogeneic [B6 → BALB/c] BMT; 9 were alive until day 44 (69%). Most of the animals succumbing ∼2 wk after allogeneic BMT showed alopecia, weight loss, hunched posture, diarrhea, and inflammation of the eyes as indicators of GvHD, but other causes like rejection, engraftment failure, infection, and rivalry were also suspected due to signs of anemia, bleeding, and skin injuries. Histology rarely showed GvHD. Only data of surviving animals were included.

A total of 100–200 μl of blood was taken from the retroorbital plexus every 5 days for 8 wk after BMT, starting on day 10. Approximately 120 days after BMT, the experiment was finished with a final single measurement.

Blood was taken from 15 BALB/c and B6 mice, each for investigation of pretransplant Ly49 expression patterns of peripheral NK cells. Nucleated cells were enriched by hypotonic lysis of erythrocytes and washing with PBS/1% BSA. After preincubation with 2.4G2 mAb and mouse serum at a dilution of 1:10 for 10 min, they were incubated for 0.5 h with the respective primary Abs at a dilution of 1:40, fixed in 1% formaldehyde, and then analyzed on a flow cytometer (Becton Dickinson FACScan, San Jose, CA). Negative controls with PBS/1% BSA instead of a primary Ab were performed for each mouse. Forward and side scatter were used to gate on the lymphocyte population. Between 5.000 and 15,000 gated events were collected for analysis. Data analysis was performed by using CellQuest (Becton Dickinson FACScan) or Windows Multiple Document Interface for Flow Cytometry (WinMDI, Scripps Research Institute, La Jolla, CA) software. The two-tailed Student’s t test for comparison of means with unequal variances was applied (Microsoft Excel software, Redmond, WA). Differences were considered significant if p < 0.05 and highly significant if p < 0.01.

Pretransplant baseline expression frequencies were calculated in two ways. For comparison of mean DX5+/Ly49+ subsets, pretransplant individual B6 and BALB/c NK cell frequencies were calculated together (see Figs. 2, 3, 5, and 6). For separate analysis of recipient or donor NK cells, means of B6 and BALB/c DX5+/Ly49+ NK cell frequencies were considered (see Fig. 4). Ly49AB6 values were referred to because mAb A1 did not detect Ly49ABALB/c.

FIGURE 2.

In vivo down-regulation of Ly49A+ NK cell frequencies after allogeneic BMT. A and C, Allogeneic [B6 → BALB/c] and syngeneic [B6 → B6] BMT of lethally irradiated mice. B and D, [BALB/c → B6] and [B6 → B6] BMT. For two-color FCM analysis, peripheral NK cells were enriched from blood samples by hypotonic lysis and labeled with anti-Ly49A-FITC (A1) and anti-DX5-PE, or anti-NK1.1-PE (PK136), a specific B6 NK cell marker. Mean values and SDs were calculated from the individual percentages of double-expressing Ly49+/DX5+ or Ly49+/NK1.1+ cells in relation to the total (single plus double stained) DX5+ or NK1.1+ NK cell population. Only B6 pretransplant values and syngeneic controls were referred to due to staining specificities of mAb A1. Note the late reduction of the recipient NK1.1+ NK cell frequency after 45 days (D) in contrast to the early decline of the donor subset (C). Syngeneic and allogeneic BMT were compared on day 55; ∗∗, p < 0.01; ∗, p < 0.05 by Student’s t test.

FIGURE 2.

In vivo down-regulation of Ly49A+ NK cell frequencies after allogeneic BMT. A and C, Allogeneic [B6 → BALB/c] and syngeneic [B6 → B6] BMT of lethally irradiated mice. B and D, [BALB/c → B6] and [B6 → B6] BMT. For two-color FCM analysis, peripheral NK cells were enriched from blood samples by hypotonic lysis and labeled with anti-Ly49A-FITC (A1) and anti-DX5-PE, or anti-NK1.1-PE (PK136), a specific B6 NK cell marker. Mean values and SDs were calculated from the individual percentages of double-expressing Ly49+/DX5+ or Ly49+/NK1.1+ cells in relation to the total (single plus double stained) DX5+ or NK1.1+ NK cell population. Only B6 pretransplant values and syngeneic controls were referred to due to staining specificities of mAb A1. Note the late reduction of the recipient NK1.1+ NK cell frequency after 45 days (D) in contrast to the early decline of the donor subset (C). Syngeneic and allogeneic BMT were compared on day 55; ∗∗, p < 0.01; ∗, p < 0.05 by Student’s t test.

Close modal
FIGURE 3.

Different modulation of donor and recipient Ly49C/I+ NK cell frequencies after allogeneic BMT. A and C, Allogeneic [BALB/c → B6] BMT and syngeneic [B6 → B6] or [BALB/c → BALB/c] BMT of lethally irradiated mice. B and D, [B6 → BALB/c] and the same syngeneic controls as in A and C. For two-color FCM analysis, peripheral NK cells were labeled with anti-Ly49C/I-PE (5E6) and anti-DX5-FITC or anti-NK1.1-FITC and analyzed as described in Fig. 2. Mean pretransplant (day 0) values of B6 and BALB/c Ly49C/I+/DX5+ subsets were calculated together. Donor or recipient B6 NK cell populations were detected by NK1.1+ mAb pre- and posttransplant for syngeneic and allogeneic BMT. Note the normalization of the donor Ly49C/IB6 NK cell frequencies in the BALB/c host but the reduction of the recipient Ly49C/IB6 after transplantation of BALB/c BM. Syngeneic and allogeneic BMT were compared on day 55; ∗∗, p < 0.01 by Student’s t test.

FIGURE 3.

Different modulation of donor and recipient Ly49C/I+ NK cell frequencies after allogeneic BMT. A and C, Allogeneic [BALB/c → B6] BMT and syngeneic [B6 → B6] or [BALB/c → BALB/c] BMT of lethally irradiated mice. B and D, [B6 → BALB/c] and the same syngeneic controls as in A and C. For two-color FCM analysis, peripheral NK cells were labeled with anti-Ly49C/I-PE (5E6) and anti-DX5-FITC or anti-NK1.1-FITC and analyzed as described in Fig. 2. Mean pretransplant (day 0) values of B6 and BALB/c Ly49C/I+/DX5+ subsets were calculated together. Donor or recipient B6 NK cell populations were detected by NK1.1+ mAb pre- and posttransplant for syngeneic and allogeneic BMT. Note the normalization of the donor Ly49C/IB6 NK cell frequencies in the BALB/c host but the reduction of the recipient Ly49C/IB6 after transplantation of BALB/c BM. Syngeneic and allogeneic BMT were compared on day 55; ∗∗, p < 0.01 by Student’s t test.

Close modal
FIGURE 5.

Comparison of Ly49+ subset frequencies between day 0 and day 55 or >120 days after allogeneic [BALB/c → B6] BMT. The mean frequencies of Ly49A+, Ly49C/I+, and double-expressing Ly49A+/Ly49C/I+ peripheral NK cell subsets of the total chimeric (donor/recipient) NK cell populations were calculated from the FCM single-expressing DX5+ and double-expressing DX5+/Ly49+ populations (∗∗, p < 0.01; ∗, p < 0.05, Student’s t test). As for DX5+/Ly49A+ and Ly49A+/Ly49C/I+, only B6 pretransplant values were referred to. Note the maintenance of all significant subset reductions 55 days and >120 days after allogeneic BMT compared with day 0.

FIGURE 5.

Comparison of Ly49+ subset frequencies between day 0 and day 55 or >120 days after allogeneic [BALB/c → B6] BMT. The mean frequencies of Ly49A+, Ly49C/I+, and double-expressing Ly49A+/Ly49C/I+ peripheral NK cell subsets of the total chimeric (donor/recipient) NK cell populations were calculated from the FCM single-expressing DX5+ and double-expressing DX5+/Ly49+ populations (∗∗, p < 0.01; ∗, p < 0.05, Student’s t test). As for DX5+/Ly49A+ and Ly49A+/Ly49C/I+, only B6 pretransplant values were referred to. Note the maintenance of all significant subset reductions 55 days and >120 days after allogeneic BMT compared with day 0.

Close modal
FIGURE 6.

Comparison of Ly49+ subset frequencies between day 0 and day 55 or >120 days after allogeneic [B6 → BALB/c] BMT. The mean frequencies of Ly49A+, Ly49C/I+, and double-expressing Ly49A+/Ly49C/I+ peripheral NK cell subsets of the total chimeric (donor/recipient) NK cell populations were calculated from the FCM single-expressing DX5+ and double-expressing DX5+/Ly49+ populations (∗∗, p < 0.01; ∗, p < 0.05, Student’s t test). As for DX5+/Ly49A+ and Ly49A+/Ly49C/I+, only B6 pretransplant values were referred to. Note the significant decrease of the Ly49C/I+ NK cell subset >120 days after allogeneic BMT compared with day 0 and in contrast to day 55. The highly significant reduction of the Ly49A+ and Ly49A+/Ly49C/I+ subset on day 55 was maintained.

FIGURE 6.

Comparison of Ly49+ subset frequencies between day 0 and day 55 or >120 days after allogeneic [B6 → BALB/c] BMT. The mean frequencies of Ly49A+, Ly49C/I+, and double-expressing Ly49A+/Ly49C/I+ peripheral NK cell subsets of the total chimeric (donor/recipient) NK cell populations were calculated from the FCM single-expressing DX5+ and double-expressing DX5+/Ly49+ populations (∗∗, p < 0.01; ∗, p < 0.05, Student’s t test). As for DX5+/Ly49A+ and Ly49A+/Ly49C/I+, only B6 pretransplant values were referred to. Note the significant decrease of the Ly49C/I+ NK cell subset >120 days after allogeneic BMT compared with day 0 and in contrast to day 55. The highly significant reduction of the Ly49A+ and Ly49A+/Ly49C/I+ subset on day 55 was maintained.

Close modal
FIGURE 4.

Separate analysis of donor and recipient Ly49C/I+ NK cell populations on day 0 and day 50 after BMT. BALB/c donor BM was transplanted into B6 recipient mice [BALB/c → B6] and B6 donor BM into BALB/c recipient mice [B6 → BALB/c]. The mean BALB/c-derived NK1.1/DX5+/Ly49+ recipient and donor NK cell populations were calculated from the absolute FCM DX5+/Ly49C/I+ and NK1.1+/Ly49C/I+ counts. BALB/c NK cells were stained with anti-DX5 and B6 NK cells with anti-NK1.1. After BMT both recipient and donor Ly49C/I+ were significantly reduced compared with day 0 (∗∗, p < 0.01; ∗, p < 0.05, Student’s t test). Note that pretransplant frequencies between both strains were similar whereas after the allogeneic BMTs a significant difference occurred in BALB/c- and B6-derived NK cells (p < 0.01, no asterisk shown).

FIGURE 4.

Separate analysis of donor and recipient Ly49C/I+ NK cell populations on day 0 and day 50 after BMT. BALB/c donor BM was transplanted into B6 recipient mice [BALB/c → B6] and B6 donor BM into BALB/c recipient mice [B6 → BALB/c]. The mean BALB/c-derived NK1.1/DX5+/Ly49+ recipient and donor NK cell populations were calculated from the absolute FCM DX5+/Ly49C/I+ and NK1.1+/Ly49C/I+ counts. BALB/c NK cells were stained with anti-DX5 and B6 NK cells with anti-NK1.1. After BMT both recipient and donor Ly49C/I+ were significantly reduced compared with day 0 (∗∗, p < 0.01; ∗, p < 0.05, Student’s t test). Note that pretransplant frequencies between both strains were similar whereas after the allogeneic BMTs a significant difference occurred in BALB/c- and B6-derived NK cells (p < 0.01, no asterisk shown).

Close modal

To test chimerism of NK cells after allogeneic BMT, two-color fluorescence staining of the total NK cell population with anti-DX5 and anti-H-2Kb or anti-H-2Kd was performed on each day of measurement. Thus engraftment success or failure was followed continuously. Syngeneically transplanted mice were used as controls.

BALB/c mice solely express the MHC class I haplotype H-2d whereas B6 mice express H-2b. Recipient or donor NK cells expressing the respective MHC class I molecule were detected by double immunofluorescence staining of the Ags H-2Kd or H-2Kb and DX5. In lethally irradiated mice of both strains, recipient NK cells were found in an initially decreasing (50–60% after 10 days, 30–40% after 20–30 days, and 20% after 30 days; data not shown), but then relatively stable proportion of 10–20% of all NK cells (Fig. 1). Donor NK cell numbers were increasing to ∼50% of the total NK cell population within the first 10 days, reaching ∼70% after 20–30 days, 80% after 30–40 days (data not shown), and 80–90% after 55 days (Fig. 1). This chimeric pattern remained stable up to 120 days, varying <10%. Hence, the Ly49 expression patterns of both recipient and donor NK cells during engraftment could be analyzed separately in mixed allogeneic BM chimeras over a long posttransplant period.

FIGURE 1.

Total chimeric (donor/recipient) DX5+ NK cell populations 55 days after allogeneic BMT. FCM panels: Transplantation of BALB/c (H2-Kd) donor BM into B6 (H2-Kb) recipient mice (top), and transplantation of B6 (H2-Kb) donor BM into BALB/c (H2-Kd) recipient mice (lower). Gating on lymphocytes in forward and sideward scatter preceded each measurement. MHC class I alleles were distinguished by anti-H2-Kb-FITC mAb and anti-H2-Kd-FITC mAb on peripheral NK cells by double-immunofluorescence labeling with anti-DX5-PE mAb, a pan-NK cell marker. Note high and low MHC class I expression in the donors. The bar chart shows mean donor (∼80–90%) and recipient (∼10–20%) proportions of the DX5/H2-Kd/b+ NK cell population as also implied in the panels above.

FIGURE 1.

Total chimeric (donor/recipient) DX5+ NK cell populations 55 days after allogeneic BMT. FCM panels: Transplantation of BALB/c (H2-Kd) donor BM into B6 (H2-Kb) recipient mice (top), and transplantation of B6 (H2-Kb) donor BM into BALB/c (H2-Kd) recipient mice (lower). Gating on lymphocytes in forward and sideward scatter preceded each measurement. MHC class I alleles were distinguished by anti-H2-Kb-FITC mAb and anti-H2-Kd-FITC mAb on peripheral NK cells by double-immunofluorescence labeling with anti-DX5-PE mAb, a pan-NK cell marker. Note high and low MHC class I expression in the donors. The bar chart shows mean donor (∼80–90%) and recipient (∼10–20%) proportions of the DX5/H2-Kd/b+ NK cell population as also implied in the panels above.

Close modal

Additionally, two populations with high and low expression of MHC class I were observed in donor but not recipient NK cells in about two-thirds of all chimeras (data not shown). Thirty to 50 days after BMT, the low-expressing population reached up to 60% of DX5+ NK cells but constituted only 10–15% of DX5+ NK cells after 120 days.

Ly49ABALB/c and Ly49AB6 both recognize H-2Dd and H-2Dk ligands, but not H-2Db ligands. Therefore, there is no known ligand for Ly49A in B6 H-2b mice. To investigate the influence of a new but cognate host environment on the expression frequency of Ly49A on H-2b donor NK cells, we transplanted B6 BM into BALB/c mice (Fig. 2, A and C). The frequencies of the Ly49A-expressing NK cell subsets after allogeneic BMT are shown together with the syngeneic B6 controls ([B6 → B6] BMT) starting from the pretransplant baseline values and ending 55 days after transplantation. The allogeneic frequencies were compared with the syngeneic on day 55. Because mAb A1 only recognized the Ly49A allele of B6 but not BALB/c mice, the Ly49A+/DX5+ NK cell population was congruent with the B6 donor-derived NK cell population. B6 specifically labeled NK1.1+ NK cells were investigated separately to check whether there was a difference compared with the results obtained with anti-DX5 (Fig. 2, C and D).

Before BMT, Ly49A was expressed on 16.5 ± 4.1% of DX5+ B6 NK cells and on 19.8 ± 3.6% of NK1.1+ B6 NK cells (p > 0.05, not significant; <1% of the gated B6 lymphocyte population was NK1.1 DX5+ in naive and syngeneically transplanted B6 mice).

As shown in Fig. 2, A and C, only the allogeneic BMT clearly resulted in altered expression frequencies of the donor Ly49A+ NK cell populations compared with the B6 syngeneic controls. After allogeneic BMT, the proportion of Ly49A-expressing NK cells declined already after ∼15–20 days post-BMT to a significantly lower level of 4.9 ± 1.7% Ly49A+/DX5+ NK cells (Fig. 2,A) and 5.4 ± 1.7% Ly49A+/NK1.1+ NK cells on day 55 (Fig. 2 C) compared with day 0 (19.8 ± 3.6%) and compared with the B6 syngeneic controls, which continued expressing the pretransplant expression pattern (day 55: 12.6 ± 1.5% Ly49A+/DX5+ NK cells; 17.2 ± 2.9% Ly49A+/NK1.1+ NK cells). Thus interaction with a new but cognate H-2d ligand of the host environment down-regulates the donor Ly49A+ H-2b NK cell frequency in peripheral blood.

It was then studied whether the exposure of recipient NK cells to a new but cognate ligand just on donor BM-derived cells, and not the whole microenvironment of the host, was sufficient to lead to a reduction of peripheral Ly49A + NK cells (Fig. 2, B and D). We performed a transplantation of BALB/c H-2d BM to a B6 H-2b background and examined the frequency of Ly49A expressing recipient-derived NK cells. Again, we found a lower frequency of Ly49A+ NK cells only after allogeneic [BALB/c → B6] BMT but not syngeneic BMT. The proportion of Ly49A+/DX5+ NK cells constantly decreased from pretransplant levels and was significantly reduced after 55 days (1.5 ± 0.7%, Fig. 2,B) compared with the B6 syngeneic controls (11.0 ± 4.8%, Fig. 2,B). The syngeneic controls kept on expressing the pretransplant expression pattern without significant changes (day 0 vs day 55: p = 0.08, Ly49A+/DX5+; p = 0.16, Ly49A+/NK1.1+). Using the NK1.1 B6 NK cell-specific marker, a significant proportional decrease of the recipient Ly49A+/NK1.1+ NK cell numbers was also observed compared with the syngeneic controls (Fig. 2 D). Interestingly, this reduction of the recipient-derived Ly49A+/NK1.1+ subset in [BALB/c → B6] BMT occurred later, after 45 days, compared with the donor-derived Ly49A+/NK1.1+ and Ly49A+/DX5+ NK cell subset in [B6 → BALB/c] BMT, which had declined early after 20 days. The reduction of the recipient DX5+ population after [BALB/c → B6] BMT was less delayed after 25 days in contrast to the NK1.1+ population. Therefore donor and recipient NK cells developed different Ly49 subset frequencies at different times.

In summary, interaction with a new but cognate H-2d donor ligand on BM-derived cells also down-regulated Ly49A+ recipient H-2b NK cell frequency effectively in a known host environment. After ∼30–40 days, the immune reconstitution of the donor-derived hematopoietic cells had taken place for the most part and ∼80% donor-derived cells were found in the peripheral blood (Fig. 1). The delayed reduction of the recipient NK cells indicated that the amount of ligand-expressing cells was relevant for this decrease.

Parallel to Ly49A, we examined the expression frequency of Ly49C/I+ NK cells. B6 NK cells express Ly49C/I whereas BALB/c NK cells only express Ly49C. Ly49CBALB/c recognizes the H-2Kb, H-2Kd, and H-2Dd ligands (36, 38). Ly49CB6 only recognizes H-2Kb. It was investigated whether the Ly49C/I+ NK cell frequency was also down-regulated upon interaction with a cognate ligand as was seen for Ly49A. The transplantation of BALB/c BM to the B6 background [BALB/c → B6] led to an interaction of the donor H-2d NK cells with a new but cognate ligand, H-2Kb, expressed in the host environment (Fig. 3, A and C). Considering our results on Ly49A-expression frequencies, a down-regulation of Ly49C-expressing BALB/c NK cell proportions could be expected.

Before BMT Ly49C/I was expressed on 37.9 ± 7.7% of DX5+ NK cells, means of both strains were calculated together. On either recipient or donor, NK1.1+ B6-derived NK cells alone Ly49C/I was expressed on 38.4 ± 6.9% NK cells. 35.8 ± 10.2% of BALB/c DX5+ NK cells expressed Ly49C (Fig. 4), revealing no statistical difference between BALB/c and B6 mice.

As shown in Fig. 3,A, the proportion of total donor/recipient Ly49C/I+/DX5+ NK cells rapidly decreased after both allogeneic [BALB/c → B6] BMT and syngeneic [BALB/c → BALB/c] BMT within the first 3 wk. In the syngeneic controls (also of [B6 → B6] BMT), they subsequently regained pretransplant frequencies (25 ± 9% on day 55) whereas after allogeneic BMT they remained significantly reduced after 55 days (8.2 ± 0.7%). The syngeneic controls exhibited no significant difference of Ly49C/I+ subsets between day 0 and day 55 (p > 0.05 B6 and BALB/c). The reduction of the total donor/recipient Ly49C/I+/DX5+ NK cell subset implied that the donor proportion within this population must have been greatly reduced as they represent ∼80% of peripheral NK cells at this time. Additional separate analysis of the donor-derived Ly49C+/NK1.1/DX5+ subset on day 50 revealed a more pronounced, statistically significant proportional reduction of the donor than of the recipient NK cells compared with day 0 (Fig. 4).

Therefore, parallel to the reduction of the Ly49A+ subset a reduction of the donor Ly49CBALB/c/H-2d NK cell frequency in a new but cognate H-2Kb-expressing host environment was also observed.

In contrast to donor Ly49CBALB/c NK cells in a H-2b host, donor Ly49C/IB6/H-2b NK cells developing in the BALB/c H-2d host encounter a noncognate ligand in the host environment because Ly49C/IB6 NK cells only recognize H-2Kb. As expected the proportion of total donor/recipient Ly49C/I+ NK cells decreased within the first 2–3 wk but then developed toward pretransplant numbers (37.9 ± 7.7%, DX5+/Ly49C/I+) 55 days after syngeneic (26 ± 7%) as well as after allogeneic (31.2 ± 3.3%) BMT (Fig. 3,B). Analyzing the Ly49C+/NK1.1+ NK cells separately confirmed the normalization for the donor subset directly, which exhibited a development of subset frequency very similar to the syngeneic controls (Fig. 3,D). Regarding the time points of variation, the early increase of the total donor/recipient Ly49C/I+/DX5+ proportion after 20–25 days to nearly pretransplant levels (Fig. 3,B) was observed similarly in the donor-derived Ly49C/I+/NK1.1+ NK cell subset (28.9 ± 2.6%) (Fig. 3 D).

This normalization of Ly49C/IB6 donor NK cell frequency in a noncognate host environment confirmed the opposite down-regulating effect of new but cognate ligands.

Confronting the recipient Ly49C/IB6 NK cells with the new H-2d donor meant interaction with a noncognate ligand because they only recognize H-2Kb. Therefore we would not expect a down-regulation. However, the recipient Ly49C/I+/NK1.1+ NK cells were also markedly reduced after 25 day up to 55 days. Again, the recipient-derived Ly49C/I+/NK1.1+ NK cells had decreased later (Fig. 3,C) compared with the constant decrease ∼10 days earlier of the total donor/recipient Ly49C/I+/DX5+ NK cell population (Fig. 3 A). Furthermore, the Ly49C/I+/NK1.1+ subset decreased earlier compared with the recipient Ly49A+/NK1.1+ subset, which was reduced after 45 days.

Thus, in contrast to the reductions upon encounter of new and cognate ligands, we observed a down-regulation of recipient Ly49C/IB6 NK cell frequency here facing a new but noncognate H-2d donor ligand in a cognate H-2b microenvironment.

In contrast to recipient Ly49C/IB6 NK cells, recipient Ly49CBALB/c/H-2d NK cells were expected to recognize the new H-2Kb donor ligand after [B6 → BALB/c] BMT. As indirectly shown in Fig. 3,B, the early reductions of the total donor/recipient Ly49C/I+/DX5+ proportions implied a reduction of the recipient subset constituting ∼40–50% of peripheral NK cells 15 days after BMT. The separate analysis of the recipient Ly49C/I+/NK1.1/DX5+ subset on day 50 showed that they remained on a significantly lower level (10.7 ± 6.9%, Fig. 4) compared with the donor-derived Ly49C/I+ subset (27.5 ± 3.2%, Fig. 4) that was also reduced but closer to pretransplant levels (38.4 ± 7.2%, Fig. 4). This difference was also implicit in the total donor/recipient Ly49C/I+/DX5+ NK cell population that showed a relatively high variation on day 50 (Fig. 3,B). Thus donor- and recipient-derived NK cells revealed statistically significant different Ly49C/I+ subsets on day 50 post-BMT (Fig. 4).

Therefore, the recipient Ly49CBALB/c/H-2d NK cells were down-regulated upon interaction with the new but cognate H-2Kb donor ligand on B6 BM-derived cells.

A final blood sample was taken from all mice after ∼120 days to compare the expression patterns with those after 55 days. As seen in the second and third columns of Figs. 5 and 6, most significant variations found 55 days after allogeneic BMT still occurred in all mice after 120 days with only minor deviations; the syngeneic controls still exhibited pretransplant frequencies (data not shown). In [BALB/c → B6] BMT, both proportional Ly49A+ and Ly49C/I+ NK cell subsets of the total donor/recipient DX5+ NK cell populations were still significantly lowered (Fig. 5). The recipient Ly49A+/NK1.1+ B6 NK cell proportions in [BALB/c → B6] BMT were less reduced (7.5 ± 6%, data not shown) than the DX5+ subset (1.2 ± 1%, Fig. 5); the Ly49C/I+/NK1.1+ recipient subset was more reduced (9.5 ± 1%, data not shown) than the total donor/recipient Ly49C/I+/DX5+ NK cell population (12.7% ± 6, Fig. 5), indicating a maintained down-regulation in the presence of the cognate H-2d donor ligand. After [B6 → BALB/c] BMT, the Ly49A+/DX5+ subset (Fig. 6) was still significantly reduced as well as the donor-derived Ly49A+/NK1.1+ subset (3.4%, Fig. 7) (5.6 ± 2.3%, data not shown) in contrast to the syngeneic controls (14.7 ± 4.1%; data not shown). The only significant difference occurring after 120 days compared with day 55 was the reduction of the previously normalized Ly49C/I+/DX5+ total donor/recipient NK cell proportion after allogeneic [B6 → BALB/c] BMT compared with day 0 (Fig. 7) and in contrast to day 55 (Fig. 6). The donor-derived Ly49C/I+/NK1.1+ subset was reduced (24.3 ± 7.7%, data not shown) (24.6%, Fig. 7) in contrast to the syngeneic controls (34.1 ± 3.3%).

FIGURE 7.

Expression of Ly49A and Ly49C/I before and >120 days after allogeneic [BALB/c → B6] and [B6 → BALB/c] BMT. Peripheral NK cells were enriched and labeled with anti-Ly49A-FITC or Ly49C/I-PE and anti-DX5-FITC/PE or anti-NK1.1-FITC/PE as described. Gating on lymphocytes in forward and sideward scatter preceded each measurement. Each of the three horizontal panels is assigned to the respective BMT and includes the Ly49+ pretransplant (day 0) total (DX5+), posttransplant total and either recipient or donor-derived NK1.1 B6 NK cell population. Pretransplant measurements of Ly49AB6, Ly49C/IB6 in [BALB/c → B6] BMT, and Ly49C BALB/c in [B6 → BALB/c] BMT were referred to. Numbers indicate the percentage of cells in the respective quadrants. Note the reduction of the total and donor Ly49C/I+ subsets (24.6% and 28.6%, respectively) in [B6 → BALB/c] BMT compared with day 0 (35.5%). The total and recipient Ly49C/I+ subsets in [BALB/c → B6] BMT are still very low (10.3% and 6.7%, respectively) compared with day 0 (38%). All Ly49A+ subsets remain decreased >120 days after BMT.

FIGURE 7.

Expression of Ly49A and Ly49C/I before and >120 days after allogeneic [BALB/c → B6] and [B6 → BALB/c] BMT. Peripheral NK cells were enriched and labeled with anti-Ly49A-FITC or Ly49C/I-PE and anti-DX5-FITC/PE or anti-NK1.1-FITC/PE as described. Gating on lymphocytes in forward and sideward scatter preceded each measurement. Each of the three horizontal panels is assigned to the respective BMT and includes the Ly49+ pretransplant (day 0) total (DX5+), posttransplant total and either recipient or donor-derived NK1.1 B6 NK cell population. Pretransplant measurements of Ly49AB6, Ly49C/IB6 in [BALB/c → B6] BMT, and Ly49C BALB/c in [B6 → BALB/c] BMT were referred to. Numbers indicate the percentage of cells in the respective quadrants. Note the reduction of the total and donor Ly49C/I+ subsets (24.6% and 28.6%, respectively) in [B6 → BALB/c] BMT compared with day 0 (35.5%). The total and recipient Ly49C/I+ subsets in [BALB/c → B6] BMT are still very low (10.3% and 6.7%, respectively) compared with day 0 (38%). All Ly49A+ subsets remain decreased >120 days after BMT.

Close modal

Double expression of Ly49A and Ly49C/I could only be examined on B6 NK cells due to the staining specificity of mAb A1. Fifty-five days after allogeneic [BALB/c → B6] BMT, recipient B6 NK cells expressing both receptors were proportionally diminished (3.4 ± 2.5%, Fig. 5). After [B6 → BALB/c] BMT, double-expressing donor B6 NK cells were also significantly reduced (4.8 ± 2.6%, Fig. 6) compared with day 0 (B6: 19 ± 6.9%, Figs. 5 and 6). Therefore 120 days after both allogeneic BMTs they were still significantly reduced (Figs. 5 and 6). There was no difference compared with the syngeneic [B6 → B6] controls as they also exhibited significantly reduced proportions 55 and 120 days after BMT (6.7 ± 3% and 8.1 ± 1%, respectively) (data not shown).

In the present study, we examined the development of Ly49A+ and Ly49C/I+ subset frequencies of peripheral NK cells over 120 days after allogeneic and syngeneic BMT. We produced fully allogeneic MHC class I-disparate BM chimeras by transplanting inbred B6 and BALB/c adult mice to study NK cell tolerance. In these chimeras, we separately analyzed recipient and donor NK cells, which constituted ∼10–20% and 80–90%, respectively, of peripheral NK cells after 40–55 days, remaining stable after ∼120 days. Because the life span of a murine NK cell is <1 wk (39), the recipient NK cells presumably consisted of radioresistant mature as well as newly developing NK cells despite the myeloablative regimen. The degree of donor chimerism and the survival rates were influenced by the number of grafted cells because transplantation of more graft cells and magnetic depletion of T cells in later experiments increased both (data not shown). Histology and blood tests provided evidence that failure of engraftment was a major cause of death rather than GvHD (data not shown).

The initially changing MHC class I environment during establishment of chimerism especially challenges the Ly49 receptor repertoire formation of the developing donor and recipient NK cells during the first 8 wk after BMT. Low expression of MHC class I could be indicative of an immature population of NK cells compared with high expression levels on mature NK cells, previously described for GM-CFU cells (40). Whether these different levels of MHC class I on NK cells influence their expression of Ly49 receptors is a question for further investigations.

During normal ontogeny after birth splenic Ly49+ NK cells gradually rise to adult numbers during the first 6–8 wk of life (41). To assure that the alterations of Ly49+ subset frequencies we observed after allogeneic BMT were not merely a result of normal ontogeneic variations of developing NK cells, we performed syngeneic controls. Pretransplant frequencies of Ly49A- and Ly49C/I-expressing NK cells were similar to other studies (42).

It has to be considered that the initial challenge for the donor NK cells entering the new MHC environment is different to later time points when the donor itself constitutes the majority of BM cells and peripheral BM-derived cells. Even then the recipient’s MHC class I expression continues to exert influence. The recipient NK cells initially face a known MHC environment but encounter an increasing number of donor cells expressing non-self MHC class I molecules. Forty to 50 days after BMT, acute rejection was most likely overcome in the surviving mice. To keep the host alive, recipient and donor NK cells must have adapted their Ly49 receptor-mediated recognition pattern by then. We hypothesize that regulation of Ly49+ subset frequencies that recognize non-self MHC class I molecules is required of donor and host NK cells alike in a chimeric host.

With the present results it was shown that regulation of Ly49+ subset frequencies is important in allogeneic BMT extending the observations of other groups that focused on expression levels in vitro (43) or in vivo after BMT (44). Our findings provide evidence that the numbers of NK cells expressing a certain Ly49 receptor are influenced by host MHC class I molecules. This influence was first described by Held et al. (33). They found higher numbers of Ly49A+ NK cells in β2m than in β2m+ mice. In the present study, we proved a regulatory effect of new MHC ligands in allogeneic BM chimeras. Above all, reductions of NK cell subset frequencies were observed with the introduction of new H-2b or H-2d ligands after allogeneic BMT. This result would be the inverse effect to the β2m mice because of more H-2 alleles being presented in the chimeric host. The reductions observed were confined to Ly49+ subsets within a total peripheral NK cell population of normal or even increased size.

Furthermore, the period of time crucial for the regulation was determined. The reductions occurred early after ∼15–20 days post-BMT in the total and donor NK cell populations. In contrast they tended to be established later after 25–45 days in the recipient subsets and were most evident after 55 days. Therefore the first 8 wk post-BMT is the critical period of time for NK cells to adapt their Ly49 expression to ensure tolerance. We assume that stability of new expression patterns depended on stability of chimerism that was reached ∼50–60 days after BMT. Ly49+ subset repertoires of mature NK cells are thought to adapt continuously (15). On the one hand, we observed rather stable frequencies of Ly49A+, Ly49C/I+, and Ly49A+/Ly49C/I+ NK cell subsets over 120 days, on the other hand minor changes of chimerism (∼10%) and of subset frequencies, e.g., the down-regulation of the previously normalized Ly49C/I+/DX5+ subset 120 days after [B6→ BALB/c] BMT. The delayed reduction of the recipient Ly49AB6 and Ly49C/I B6 subsets might be explained by the initially lower but then growing impact of graft MHC reaching full BM extension 40–50 days after BMT. In contrast, the earlier reduction of the donor Ly49AB6 and Ly49C/I B6 NK cell proportions 15–20 days after [B6 → BALB/c] BMT could result from the greater impact of the external non-self recipient BALB/c environment from the beginning of engraftment, including the decreasing irradiated hematopoietic cells as well as nonhematopoietic elements. Accordingly, the strong decrease of the total Ly49C/I+/DX5+ subset on day 15 after [BALB/c → B6] BMT also implied a reduction of the donor Ly49CBALB/c subset.

To interpret these results we refer to the calibration model describing down-regulating effects on Ly49 expression levels by MHC-ligands in the surrounding external environment (14) and as recently proposed also for internal MHC expressed by NK cells themselves (15, 22). In the present study, it was investigated whether this model also applied to calibration of NK cell frequencies and whether the influence by external ligands was different when expressed in the whole microenvironment of the recipient or only by donor BM cells and BM-derived donor cells. In our chimeric models, both recipient and donor MHC class I ligands obviously had a down-regulating influence on Ly49+ NK cell population sizes as discussed below.

First we examined the influence of a new but cognate host environment on the expression frequency of Ly49A B6/H-2b donor-derived NK cells in [B6 → BALB/c] BMT. The donor Ly49AB6 NK cell proportions were significantly reduced. Ly49AB6 NK cells recognize H-2Dd and H-2Dk ligands (38). Thus they recognized the external non-self H-2Dd ligand expressed by the recipient BALB/c environment, a new but cognate ligand, and were therefore down-regulated. Then we examined the influence of donor BM cells and BM-derived donor cells on recipient Ly49AB6 NK cells in [BALB/c → B6] BMT. We also observed a reduction of the recipient Ly49AB6 NK cells. They recognized the external non-self donor H-2Dd ligand expressed by the graft in the chimeric BM and by BM-derived donor cells in the periphery, a new but cognate ligand, and were therefore down-regulated as well.

As for Ly49C/I we examined the modulation of donor Ly49CBALB/c/H-2 d NK cell proportions in the H-2 b host environment after [BALB/c → B6] BMT and found a significant reduction. Ly49CBALB/c recognizes the ligands H-2Kb, H-2Kd and H-2Dd. Thus the donor Ly49CBALB/c NK cells in [BALB/c → B6] BMT can interact with the cognate H-2Kb ligand of the host environment resulting in down-regulation. Vice versa we examined the modulation of the recipient Ly49CBALB/c subset by the H-2b ligand. Here also, a decrease of the recipient Ly49CBALB/c subset was observed 50 days after [B6→ BALB/c] BMT and was implied in the reduced mean total Ly49C/I+/DX5+ subset frequency after 120 days. This is likely to be due to interaction with the donor-derived cognate H-2Kb ligand. In contrast to the donor Ly49CBALB/c NK cells, donor Ly49C/IB6/H-2b NK cells develop in a new and noncognate H-2d host environment after the [B6 → BALB/c] BMT because they only recognize H-2Kb. As they did not recognize the external non-self H-2d ligand they were not down-regulated. Another important aspect is that this increase of Ly49CB6 NK cells could also foster the engraftment of their own BM cells as described in vitro by Murphy et al. (18) for syngeneic Ly49C+ donor NK cells compared with Ly49C NK cells. Herein may lie an explanation for the repeatedly better survival rate of H-2b → H-2d bone chimeras in contrast to H-2d → H-2b chimeras. Yet, an increase of the Ly49I+ proportion alone along with a decreased Ly49C+ subset has to be taken into account. The analysis of the double-expressing Ly49A+ and Ly49C/I+ NK cells revealed a significant decrease 55 and 120 days after both allogeneic and syngeneic BMTs. The mechanisms of this similar alteration remain unclear at present.

Regarding the maintenance of the altered expression patterns 120 days after [B6 → BALB/c] BMT compared with 55 days, the previously normalized total donor/recipient Ly49C/I+/DX5+ subset frequency had significantly declined implying a reduction also of the Ly49CB6 donor subset. This reduction might indicate a down-regulating influence of the donor-derived external and/or even internal self H-2b ligands which is presumably not strong enough within the first 8 wk but increases with the growing graft and leads to more interaction of donor Ly49CB6 NK cells with H-2Kb on surrounding cells. This reduction also occurred much later in contrast to the early reduction of the recipient Ly49C/IB6 NK cells after transplantation of the H-2d noncognate ligand in [BALB/c → B6] BMT. There, the recipient Ly49C/IB6 NK cells were reduced in the presence of the non-self noncognate donor H-2d ligand and the self cognate H-2Kb ligand. This might be explained by the initially greater amount of surrounding self ligand-expressing cells in contrast to the late reduction of the donor Ly49CB6 subset above. This observation is also consistent with the delayed reduction of the recipient Ly49AB6 NK cells.

We cannot clearly define the influence of the noncognate donor H-2d ligand and the self cognate H-2Kb ligand, but postulate that the Ly49CB6 recipient NK cells were only reduced because 1) they express the cognate ligand themselves, and/or 2) they recognize it on surrounding cells expressing the self cognate H-2Kb ligand, and 3) this effect only occurs in chimeric but not syngeneically transplanted hosts. This result might provide the first in vivo observation of internal calibration with respect to the frequency of Ly49-expressing NK cells or a regulatory effect of self external ligands only in chimeric but not syngeneic hosts. A down-regulatory effect of the noncognate donor H-2d ligand is unlikely because 1) it is simply not recognized and 2) our findings primarily showed that reduction of Ly49A+ and Ly49C/I+ subsets occurred upon interaction with cognate ligands. The opposite normalization of the donor Ly49CB6 NK cell frequency in the noncognate H-2d host confirmed this assumption. The influence of Ly49IB6 that does not bind to H-2d or H-2b target cells or any other known ligand (36, 38) remains unclear. Therefore our observations provide further evidence in line with a recent study (34) that self-MHC ligands also down-regulate Ly49+ subsets of MHC-identical NK cells only in allogeneic but not in syngeneic hosts. Manilay et al. (34) showed that numbers of donor and host Ly49A+ NK cells do not change in allogeneic H-2d → H-2b BM chimeras and donor Ly49CBALB/C NK cells were first increased then the same as those of nontransplanted controls whereas recipient NK cells were decreased after 7 wk.

We conclude from our results that external cognate non-self MHC class I ligands down-regulate the frequencies of developing MHC-different Ly49+ NK cells in a chimeric host. Furthermore we assume that only in a chimeric host cognate self external ligands are also effective in the absence of a cognate non-self ligand. We also confirm that alterations are determined by the amount of ligands dependent on population sizes. During their development in the same host MHC class I-different recipient and donor NK cells adapted different Ly49 expression patterns within the first 55 days but after 120 days they both had acquired reduced Ly49+ subset frequencies. Therefore the frequencies between the Ly49A and Ly49C/I subsets differ, but are balanced between the strains as in nontransplanted MHC class I-different mice. In conclusion these results suggest a bilateral process of adaptation in mixed BM chimeras and strengthen the notion that maintenance of self-tolerance by NK cells in a changing, heterogeneous MHC environment is important for successful engraftment.

1

This study was supported by the Deutsche Forschungsgemeinschaft, SFB265 project B1.

3

Abbreviations used in this paper: BM, bone marrow; β2m, β2-microglobulin; KIR, killing inhibitory receptor; BMT, BM transplantation; GvHD, graft-vs-host disease; FCM, flow cytometry.

1
Yu, Y. Y. L., V. Kumar, M. Bennett.
1992
. Murine natural killer cells and marrow graft rejection.
Annu. Rev. Immunol.
10
:
189
2
George, T., Y. Y. L. Yu, J. Liu, C. Davenport, S. Lemieux, E. Stoneman, P. A. Matthew, V. Kumar, M. Bennett.
1997
. Allorecognition by murine natural killer cells: lysis of T-lymphoblasts and rejection of bone marrow grafts.
Immunol. Rev.
155
:
29
3
Bordignon, C., J. P. Daley, I. Nakamura.
1985
. Hematopoietic histoincompatibility reactions by NK cells in vitro: model for genetic resistance to marrow grafts.
Science
230
:
1398
4
Chadwick, B. S., R. G. Miller.
1992
. Hybrid resistance in vitro. Possible role of both class I MHC and self peptides in determining the level of target cell sensitivity.
J. Immunol.
148
:
2307
5
Bix, M., N.-S. Liao, M. Zijlstra, J. Loring, R. Jaenisch, D. Raulet.
1991
. Rejection of class I MHC-deficient hematopoietic cells by irradiated MHC-matched mice.
Nature
349
:
329
6
Liao, N.-S., M. Bix, M. Zijlstra, R. Jaenisch, D. Raulet.
1991
. MHC class I deficiency: susceptibility to natural killer (NK) cells and impaired NK activity.
Science
253
:
199
7
Meyer, D., J. Schubert, C. Scholz, S. Braun, R. E. Schmidt.
1997
. MHC-I antigen expression determines sensitivity of hematopoietic progenitor cells as targets for NK cells.
Immunobiology
197
:
494
8
Raulet, D. H..
1993
. MHC class I-deficient mice.
Adv. Immunol.
55
:
381
9
Ohlen, C., G. Kling, P. Höglund, M. Hansson, G. Scangos, C. Bieberich, G. Jay, K. Kärre.
1989
. Prevention of allogeneic bone marrow graft rejection by H-2 transgene in donor mice.
Science
246
:
666
10
Olsson, M. Y., K. Kärre, C. Sentman.
1995
. Altered phenotype and function of natural killer cells expressing the major histocompatibility complex receptor Ly-49 in mice transgenic for its ligand.
Proc. Natl. Acad. Sci. USA
92
:
1649
11
Cambiaggi, A., C. Verthuy, P. Naquet, F. Romagné, P. Ferrier, R. Biassoni, A. Moretta, L. Moretta, E. Vivier.
1997
. Natural killer cell acceptance of H-2 mismatch bone marrow grafts in transgenic mice expressing HLA-Cw3 specific killer cell inhibitory receptor (CD158b).
Proc. Natl. Acad. Sci. USA
94
:
8088
12
Kung, S. K. P., R.-C. Su, J. J. K. Graham, J. W. Chamberlain, R. G. Miller.
1998
. NK cells from human MHC class I (HLA-B) transgenic mice do not mediate hybrid resistance killing against parental nontransgenic cells.
J. Immunol.
160
:
674
13
Höglund, P., J. Sundbäck, M. Y. Olsson-Alheim, M. Johansson, M. Salcedo, H.-G. C. Öhlén, C. L. Sentman Ljunggren, K. Kärre.
1997
. Host MHC class I gene control of NK cell specificity in the mouse.
Immunol. Rev.
155
:
11
14
Sentman, C. L., M. Y. Olsson, K. Kärre.
1995
. Missing self recognition by natural killer cells in MHC class I transgenic mice. A “receptor calibration” model for how effector cells adapt to self.
Semin. Immunol.
7
:
109
15
Kåse, A., M. H. Johannson, M. Y. Olsson-Alheim, K. Kärre, P. Höglund.
1998
. External and internal calibration of the MHC class I-specific receptor Ly49A on murine natural killer cells.
J. Immunol.
161
:
6133
16
Murphy, W. J., J. R. Keller, C. L. Harrison, H. A. Young, D. L. Longo.
1992
. Interleukin-2-activated natural killer cells can support hematopoiesis in vitro and promote marrow engraftment in vivo.
Blood
80
:
670
17
Wong, S., J. D. Freeman, C. Kelleher, D. Mager, F. Takei.
1991
. Ly-49 multigene family. New members of a superfamily of type II membrane proteins with lectin-like domains.
J. Immunol.
147
:
1417
18
Murphy, W. J., A. Raziuddin, L. Mason, V. Kumar, M. Bennett, D. L. Longo.
1995
. NK cell subsets in the regulation of murine hematopoiesis. I. 5E6+ NK cells promote hematopoietic growth in H-2d strain mice.
J. Immunol.
155
:
2911
19
McQueen, K. L., J. D. Freeman, F. Takei, D. L. Mager.
1998
. Localization of five new Ly49 genes, including three closely related to Ly49C.
Immunogenetics
48
:
174
20
Mason, L. H., S. K. Anderson, W. M. Yokoyama, H. R. C. Smith, R. Winkler-Pickett, J. R. Ortaldo.
1996
. The Ly-49D receptor activates murine natural killer cells.
J. Exp. Med.
184
:
2119
21
Smith, M. K., J. Wu, A. B. H. Bakker, J. H. Phillips, L. L. Lanier.
1998
. Cutting edge: Ly-49D and Ly-49H associate with DAP12 and form activating receptors.
J. Immunol.
161
:
7
22
Andersson, M., S. Freland, M. H. Johansson, R. Wallin, J. K. Sandberg, B. J. Chambers, B. Christensson, U. Lendahl, S. Lemieux, M. Salcedo, H.-G. Ljunggren.
1998
. MHC class I mosaic mice reveal insights into control of Ly49C inhibitory receptor expression in NK cells.
J. Immunol.
161
:
6475
23
Yu, Y. Y. L., T. George, J. R. Dorfman, J. Roland, V. Kumar, M. Bennett.
1996
. The role of Ly49A and 5E6 (Ly49C) molecules in hybrid resistance mediated by murine natural killer cells against normal T cell blasts.
Immunity
4
:
67
24
Brennan, J., D. Mager, W. Jefferies, F. Takei.
1994
. Expression of different members of the Ly-49 gene family defines distinct natural killer cell subsets and cell adhesion properties.
J. Exp. Med.
180
:
2287
25
Freland, S., B. J. Chambers, M. Andersson, L. V. Kaer, H.-G. Ljunggren.
1998
. Rejection of allogeneic and syngeneic but not MHC class-I deficient tumor grafts by MHC class I-deficient mice.
J. Immunol.
160
:
572
26
Höglund, P., E. C. Öhlén, L. Carbone, H. G. Franksson, A. Ljunggren, B. Koller Latour, K. Kärre.
1991
. Recognition of β2-microglobulin-negative (β2m) T-cell blasts by natural killer cells from normal but not from β2m mice: nonresponsiveness controlled by β2m bone marrow in chimeric mice.
Proc. Natl. Acad. Sci. USA
88
:
10332
27
Lanier, L. L..
1998
. NK cell receptors.
Annu. Rev. Immunol.
16
:
359
28
Sundbäck, J., K. Kärre, C. L. Sentman.
1996
. Cloning of minimally divergent allelic forms of the natural killer (NK) receptor Ly-49C, differentially controlled by host genes in the MHC and NK gene complex.
J. Immunol.
157
:
3936
29
Held, W., B. Kunz.
1998
. An allele-specific, stochastic gene expression process controls the expression of multiple Ly49 family genes and generates a diverse, MHC-specific NK cell receptor repertoire.
Eur. J. Immunol.
28
:
2407
30
Raulet, D. H., W. Held, I. Correa, J. Dorfman, M.-F. Wu, L. Corral.
1997
. Specificity, tolerance and developmental regulation of natural killer cells defined by expression of class I specific Ly49 receptors.
Immunol. Rev.
155
:
41
31
Dorfman, J. R., D. H. Raulet.
1996
. Major histocompatibility complex genes determine natural killer cell tolerance.
Eur. J. Immunol.
26
:
151
32
Karlhofer, F. M., R. Hunziker, A. Reichlin, D. H. Margulies, W. M. Yokoyama.
1994
. Host MHC class I molecules modulate in vivo expression of a NK cell receptor.
J. Immunol.
153
:
2407
33
Held, W., J. R. Dorfman, M. F. Wu, D. H. Raulet.
1996
. Major histocompatibility complex class I dependent skewing of the natural killer cell Ly49 receptor repertoire.
Eur. J. Immunol.
26
:
2286
34
Manilay, J. O., G. L. Waneck, M. Sykes.
1998
. Altered expression of Ly-49 receptors on NK cells developing in mixed allogeneic bone marrow chimeras.
Int. Immunol.
10
:
1943
35
Rolstad, B., W. E. Seaman.
1998
. Natural killer cells and recognition of MHC class I molecules: new perspectives and challenges in immunology.
Scand. J. Immunol.
47
:
412
36
Brennan, J., S. Lemieux, J. D. Freeman, D. L. Mager, F. Takei.
1996
. Heterogeneity among Ly-49C natural killer (NK) cells: characterization of highly related receptors with differing functions and expression patterns.
J. Exp. Med.
184
:
2085
37
Cudkowitz, G..
1980
. Natural resistance to foreign hematopoietic and leukemia grafts. G. Cudkowitz, and M. Landy, and G. Shearer, eds.
Natural Resistance Systems Against Foreign Cells, Tumors and Microbes
3
Academic, New York.
38
Brennan, J., G. Mahon, D. L. Mager, W. A. Jefferies, F. Takei.
1996
. Recognition of class I major histocompatibility complex molecules by Ly-49: specificities and domain interactions.
J. Exp. Med.
183
:
1553
39
Miller, S. C..
1982
. Production and renewal of murine natural killer cells in the spleen and bone marrow.
J. Immunol.
129
:
2282
40
Griffin, J. D., K. D. Sabbath, F. Herrmann, P. Larcom, K. Nichols, M. Kornacki, H. Levine, S. A. Cannistra.
1985
. Differential expression of HLA-DR antigens in subsets of human CFU-GM.
Blood
66
:
788
41
Dorfman, J. R., D. H. Raulet.
1998
. Acquisition of Ly49 receptor expression by developing natural killer cells.
J. Exp. Med.
187
:
609
42
Sentman, C. L., J. Hackett, Jr, V. Kumar, M. Bennett.
1989
. Identification of a subset of murine natural killer cells that mediates rejection of Hh-1d but not Hh-1b bone marrow grafts.
J. Exp. Med.
170
:
191
43
Salcedo, M., M. Andersson, S. Lemieux, L. Van Kaer, B. J. Chambers, H.-G. Ljunggren.
1998
. Fine tuning of natural killer cell specificity and maintenance of self tolerance in MHC class I-deficient mice.
Eur. J. Immunol.
28
:
1315
44
Salcedo, M., A. D. Diehl, M. Y. Olsson-Alheim, J. Sundbäck, L. V. Kaer, K. Kärre, H.-G. Ljunggren.
1997
. Altered expression of Ly49 inhibitory receptors on natural killer cells from MHC class I-deficient mice.
J. Immunol.
158
:
3174