Male C.B10-H2^b/LiMcdJ (BALB.B) and female C57BL/6 (B6) mice were obtained from the Jackson Laboratory (Bar Harbor, ME, USA) and used as recipients and donor of bone marrow (BM) and spleen cells, respectively. Mice were housed under specific pathogen-free conditions at the center for animal resource development of Seoul National University College of Medicine in Korea. All the experiments were performed under approval from The Seoul National University Institutional Animal Care and Use Committee (IACUC).

Bone marrow (BM) cells were prepared from female B6 mice by flushing femurs and tibias with 1×PBS (Phosphate buffered saline). Spleen cells from female B6 mice were prepared by grinding spleens through steel mesh. Recipient male BALB.B mice were irradiated with split does from a ¹³⁷Cs source with a 5 hr interval. The recipient mice pre-conditioned either with 900 cGy (myeloablative conditioning) or 400 cGy (non-myeloablative conditioning) were injected with a mixture of B6 BM (5×10⁶) and splenocytes (2×10⁷) in a volume of 300 µL into a lateral tail vein after 5 hours after the second irradiation.

BALB.B GVHD hosts were perfused with PBS/heparin (75 U/ml; Sigmal, St Louis, MO, USA) before sacrifice to deplete circulating peripheral blood leukocytes (PBLs). The preparation of leukocytes infiltrating livers and lungs was described in detail previously (17). In brief, livers were pressed through stainless steel mesh and suspended in 5% fetal calf serum (FCS)-PBS. The cell suspensions were admixed with 33% Percoll containing 100 U/ml heparin and were then centriguged 200 rpm for 15 min at room temperature. The cell suspensions were applied to Percoll gradients and RBC lysis, and then resuspended in 5% FCS-PBS. Lungs were cut into small pieces and treated with collagenase at 37℃ for 30 minutes after washing with 1×Hanks balanced salt solution (HBSS)/HEPES(N-2-hydroxyethylpiperzine-N'-2-ethaaaanesulfonic acid: 10 mM). Then, after washing twice with 5% FCS-PBS, the cells were applied to Percoll gradients and resuspended in 5% FCS-PBS after washing with 1×PBS.

Fresh PBLs, splenocytes, or lymphocytes infiltrating livers and lungs from the GVHD hosts were incubated at 4℃ for 30 minutes in FACs buffer (1×PBS with 0.1% bovine calf serum and 0.05% sodium azide) containing PE-labeled H60/H-K^b tetramer and saturating amount of FITC-conjugated anti-CD11a (2D7, eBiosicence, San Diego, CA) and APC-conjugated anti-CD8 (eBioscience) monoclonal antibodies. Anti-β₂ microglobulin B (anti-β₂m^b; S19.8, Santa Cruz Biotechnology) antibody was used to distinguish donor (B6:β₂m^b)-originated cells from recipient (BALB.B: β₂m^a)-originated ones. Other monoclonal antibodies used were PE or APC-conjugated anti-Mac-1 (M1/70, eBioscience), FITC-conjugated anti-Gr-1 (RB6-8C5, eBioscience), PE-conjugated anti-CD4 (GK1.5, eBioscience). For intracytoplsmic staining for IFN-γ and IL-17 production, cells were stained with anti-Mac-1 or anti-Gr-1 antibodies, washed, and then fixed with 4% paraformaldehyde at room temperate for 20 minutes. After washing with 1×PBS, the cells were permeabilized with saporine and stained with APC-conjugated anti-IFN-γ (XMG1.2, eBioscience) or PE-conjugated IL-17 (IL-17A; TC11-18H10, BD Bioscience, San Diego, CA) antibodies at 4℃ for 2 hrs. The stained cells were analyzed using a FACSCalibur equipped with CellQeust software (BD Bioscience).

To select two different irradiation doses that correspond to myeloablative high intensity vs non-myeloablative mild intensity doses in the B6 → BALB.B GVHD system, 900 cGy, 600 cGy, and 400 cGy of irradiation doses were tested for chimerism formation in the PBLs of BALB.B host mice. BALB.B mice were irradiated with the three different doses, transplanted with BM and splenocytes from B6 mice, and then eye-bled on days 4, 7, and 14 post-transplantation. Pooled PBLs prepared from three BALB.B mice of each group were stained with antibodies for marker antigens, Gr-1, Mac-1, CD4 and CD8, in combination with anti-β2m^b antibody to distinguish the cell origins, donor (β2m^b) or recipient (β2m^a), and then analyzed via flow cytometry. Flow cytometry results showed that, on day 4 after the transplantation with 900 cGy irradiation dose, Gr-1⁺, Mac-1⁺, CD4⁺ T cell and CD8⁺ T cells in PBL of BALB.B hosts showed mixed chimerism, mixed with cells of donor and recipient origins. However, from day 7, they showed full chimerism with all the cell populations replaced with the cells of B6 origin (Fig. 1). On the contrary, host-originated Gr-1⁺ and Mac-1⁺ populations were still detected in of the PBLs of the BALB.B GVHD hosts with irradiation dose of 400 cGy, showing mixed chimerism status, until the day 14 after the transplantation, while most of CD4 T and CD8 T cells were of B6 origin, probably due to expansion of allo-reactive CD4 and CD8 T cells of B6 origin in the BALB.B hosts. With 600 cGy of irradiation dose, overall pattern of PBL chimerism was similar to that seen in BALB.B recipients with 900 cGy irradiation (data not shown). Based upon these results, we selected 900 cGy and 400 cGy irradiation doses as myeloablative and non-myeloablative conditioning doses, leading to full and mixed-chiemerism in the PBLs of the BALB GVHD hosts, respectively.

Next, we examined the effect of the selected myeloablative (900 cGy) and non-myeloablative (400 cGy) irradiation doses on GVHD severity in our allogeneic B6 → BALB.B model. BALB.B mice which were irradiated under the two different conditions and transplanted with allogeneic B6 BM and spleen cells were observed for their survival for >50 days and regularly checked for their weight changes for 35 days after transplantation with BM and splencotyes from B6 mice. Syngeneic control transplantation, B6 → B6, was performed under the same irradiation conditions. Myeloablative high dose pre-conditioning and subsequent allogeneic transplantation led to death of BALB.B hosts with drastic weight loss and other GVHD symptoms, such as diarrhea and hunched posture (Fig. 2). With non-myeloablative low dose pre-conditioning, allogeneic BALB.B hosts survived longer than 50 days. None of the syngeneic B6 hosts developed GVHD under any circumstatnces. Even though the B6 hosts under the myeloablative condition went though weight loss early after transplantation, they recovered soon afterward and survived longer. This result demonstrated that the selected irradiation does of 900 cGy (termed myeloablative) and 400 cGy (non-myeloablative) resulted in different outcome of GVHD induction, not only different PBL chimerism in B6 → BALB.B system, implying that they are suitable doses for investigating the irradiation effects on immune kinetics during GVHD development.

Then, we compared PBL profiles along the time course after GVHD induction between the BALB.B hosts conditioned with myeloabltive vs non-myeloablative irradiation, wondering whether PBL profiles would be different depending upon the GVHD severity, so that they could be used to distinguish and predict GVHD outcomes. Bloods from the GVHD survivors of each pre-conditioning group were collected periodically and pooled for PBL preparation. The PBLs were typed for profiling proportions of Gr-1⁺, Mac-1⁺, CD4⁺ T and CD8⁺ T cells in the blood via staining with antibodies specific for corresponding antigens and subsequent flow cytometric analysis, and the proportions of each cell type were plotted as kinetics graphs. On the contrary to our expectation, there was no big difference in the percentages of Gr-1⁺ or Mac-1⁺ cells present in bloods between the myeloabaltively vs non-myeloablatively conditioned BALB.B hosts. Their fractions reached peak, about 90% of PBL, on day 7 after the GVHD induction and continued to be high afterward (Fig. 3). The kinetics of CD4⁺ T cells in the blood was not significantly different between the two groups, either. The only significant difference was found in CD8 T cell population: the proportion of CD8 T cells in bloods of GVHD hosts was higher under the myeloablative (~12%) condition than the non-myeloabative (4~7%) condition (Fig. 3). The results from the PBL profile study demonstrated that there was no big difference in myeloid cell proportions between the GVHD hosts with myeloablative vs non-myeloablative irradiation, even though the GVHD hosts with the two different pre-conditioning doses experienced different degrees of the disease severity.

Since the PBLs profiles of GVHD hosts with different doses of irradiation were not significantly different, especially considering the proportions of myeloid lineage cells, we explored the leukocytes that infiltrated into target organs during GVHD course, presuming that the infiltration of myeloid cells into target tissues might have been influenced by the irradiation doses. Leukocytes infiltrating spleens, livers and lungs were periodically prepared from three BALB.B mice under each condition, pooled for staining and subsequent flow cytometry to profile Gr-1⁺, Mac-1⁺, CD4⁺ T and CD8⁺ T cells in the organs along the GVHD course.

Overall kinetics of the four types of leukocytes infiltrating target tissues was significantly different, depending upon the pre-conditioning doses (Fig. 4). Proportions of Gr-1⁺ cells continuously increased in the spleens (Fig. 4A), livers (Fig. 4B), and lungs (Fig. 4C) of BALB.B hosts with high intensity irradiation and accounted for more than 60% of filtrating leukocytes at late time points (later than day 21 post-transplantation) in the livers and lungs. With low intensity irradiation, the kinetics of Gr-1⁺ infiltrating cells showed patterns of early increase followed by decrease in the spleens, livers, and lungs of the BALB.B hosts. The kinetics of Mac-1⁺ cells was very similar to that of Gr-1⁺ cells residing in all the three organs in both myeloabltively and non-myeloabltively conditioned hosts.

Percentages of CD4⁺ T cells reached peak on day 4 (17%), day 7 (45%), and day 14 (32%) in the spleens, livers, and lungs of the BALB.B hosts with high intensity irradiation, respectively, and then waned afterward (Fig. 4A-C). This initial up and then down pattern of kinetics was observed in the case of CD4 T cells infiltrating the tissues of the BALB.B hosts with low intensity irradiation, but the peak values were comparatively low (14% in spleens, 30 % in livers, and 10% in lungs on day 4 post-transplantation).

Patterns of CD8 T cell kinetics was similar to those of CD4 T cell in the tissues of the GVHD hosts with myeloablative pre-conditioning with peak values of 13% (day 4 post-transplantation), 35% (day 14), and 42% (day 14) in the spleens, livers, and lungs, respectively. In the GVHD hosts with non-myeloablative pre-conditioning, proportion CD8 T cells gradually increased in the spleens with peak values of 12% on day 21 but, in the lungs and livers, they reached peak on day 14 post-transplantation.

Altogether the results obtained from the profiling of leukocytes infiltrating target organs of GVHD, it could be concluded that immune kinetics in the target organs is influenced by the pre-conditioning intensity. Among the three organs, liver and lung showed very similar kinetics, while spleen showed its own kinetics pattern for each cell population, possibly due to its role as secondary lymphoid organ as well as a GVHD target organ. However, on the contrary to our expectation, proportion of myeloid cells infiltrating spleens and lungs cells at early time point (on day 4 post-transplantation) was lower in the GVHD hosts with myeloabaltive irradiation than non-myeloablative irradiation, for example, 12 % vs 15 % of Gr-1⁺ cells in the spleens and 17.3% vs 44% of Gr-1⁺ cells in the lungs of myeloabaltively conditioned vs non-myeloabaltively conditioned hosts, respectively.

Even though proportions of myeloid populations in target tissues on day 4 post-transplantation did not correlate with irradiation intensities, we wondered whether proportion of cells producing inflammatory cytokine in the infiltrating myeloid cells would correlate with the irradiation dose. To test this, we prepared infiltrating leukocytes from the spleen, liver, and lung on day 4 after GVHD induction and analyzed for the production of IL-17 by Gr-1⁺ or Mac-1⁺ cells. Intracytoplasmic staining of the cells with anti-IL-17 and anti-IFN-γ antibodies and subsequence flow cytometry analysis revealed that proportions of IL-17 producing cells in Gr-1⁺ and Mac-1⁺ cells present in the spleen were higher when the hosts received myeloabaltively high dose irradiation than those with non-myeloabaltively low dose irradiation (Fig. 5A). They did not produce IFN-γ. Proportion of IL-17 producing Gr-1⁺ cells was higher in the spleens (18.7% on average) than in the livers (7.63% on average) or lungs (10.4% on average) of the GVHD hosts with myeloablative irradiation (Fig. 5B). In the case of Mac-1⁺ cells, proportions of IL-17 producing cells were higher in the lungs (16.7% on average) than those in the spleens (5.85% on average) and the livers (5.13% on average) of the myeloablatively conditioned hosts. In all the three organs, proportions of IL-17 producing cells were lower in non-myeloablatively conditioned hosts than in the myeloabltively conditioned ones. On day 7 after the GVHD induction, IL-17 production by the myeloid cells that infiltrated target organs subsided to very low levels (Fig. 5B). Altogether, these results demonstrate that, even though proportions of Gr-1⁺ or Mac-1⁺ cells was little lower in the target organs of myeloablatively conditioned hosts than those of non-myeloabaltively conditioned ones at early time points after GVHD induction, proportions of myeloid cells producing inflammatory cytokine IL-17 in the infiltrating cells were higher in the former than the latter.