We have examined how the host environment influences the graft-vs-leukemia (GVL) response following transfer of donor T cells to allogeneic chimeras. Donor T cells induce significant GVL when administered in large numbers to established mixed chimeras (MC). However, when using limiting numbers of T cells, we found that late transfer to MC induced less GVL than did early transfer to freshly irradiated allogeneic recipients. Late donor T cell transfer to MC was associated with marked accumulation of anti-host CD8 cells within the spleen, but delayed kinetics of differentiation, reduced expression of effector molecules including IFN-γ, impaired cytotoxicity, and higher rates of sustained apoptosis. Furthermore, in contrast to the spleen, we observed a significant delay in donor CD8 cell recruitment to the bone marrow, a key location for hematopoietic tumors. Increasing the numbers of T cells transferred to MC led to the enhancement of CTL activity and detectable increases in absolute numbers of IFN-γ+ cells without inducing graft-vs-host disease (GVHD). TLR-induced systemic inflammation accelerated differentiation of functional CTL in MC but was associated with severe GVHD. In the absence of inflammation, both recipient T and non-T cell populations impeded the full development of GVHD-inducing effector function. We conclude that per-cell deficits in the function of donor CD8 cells activated in MC may be overcome by transferring larger numbers of T cells without inducing GVHD.

Following allogeneic bone marrow transplantation (BMT),5 donor T cell alloreactivity can be exploited to generate a powerful anti-leukemia effect, a phenomenon termed the graft-vs-leukemia (GVL) response (1, 2). Unfortunately, under conditions in which donor T cells are transferred immediately to lethally irradiated recipients, GVL responses are often associated with the development of graft-vs-host disease (GVHD). Delaying the timing of donor T cell administration by 2–3 wk may reduce the risk of GVHD in full allogeneic chimeras and still permit the induction of GVL, but by 4–5 wk, graft-vs-host (GVH) reactivity cannot be induced (3, 4). In sharp contrast, the GVL response remains intact when donor T cells are infused (at >8 wk) into established mixed chimeras (MC), where hematopoietic elements from both the donor and recipient are present at the time of transfer (5, 6, 7). This finding relates to the continued presence of large numbers of host hematopoietic APCs, which are required for maximal priming of donor T cells recognizing recipient class I (7) and class II (5) following MHC-mismatched transplantation.

Although high numbers of donor T cells activated in MC do not cause GVHD, they do so readily upon transfer to secondary, irradiated allogeneic recipients (8). Moreover, the application of a local or systemic TLR stimulus allows such donor T cells to cause local or systemic GVHD, respectively (8). These studies indicate that donor leukocyte infusion (DLI)-derived GVH-reactive T cell populations activated in MC have no absolute, intrinsic functional defect. Rather, these and other data (9, 10, 11, 12, 13, 14) argue that extrinsic factors such as inflammation within the host environment are critical for the recruitment of activated T cells to the epithelial target tissues and hence the development of GVHD. It is still not known, however, whether anti-host CTL arising in established MC or freshly irradiated allogeneic (TBI-allo) recipients are fully equivalent in terms of functional activity and their capacity to induce GVL. Potentially important differences between the two host environments include the duration of direct Ag presentation by professional APC, the levels of lymphopenia-induced proliferation, the extent of suppression mediated by regulatory cell populations, and the degree of inflammation present at the time of T cell transfer. These and other factors have the potential to influence the precise cellular mechanisms of GVL operating in the two contexts. In support of this notion, CD4 cell help is not required for CD8 cell-mediated GVL following immediate transfer of donor T cells but is an absolute requirement following delayed transfer to MC (5). Better definition of the properties of GVH-reactive CD8 cells emerging following delayed T cell transfer to MC will be important in the development of strategies that seek to use DLI to induce GVL in the clinical setting.

The reduction in GVHD observed following delayed as compared with immediate DLI to full chimeras has been well documented (3, 4, 15). However, in this study we wished to examine the functional properties of GVH-reactive CD8 cells following delayed transfer to MC (where priming host APCs are present in large numbers) and compare them to those observed following immediate transfer to TBI-allo recipients. Using a tumor protection model that measures a donor CD8 cell-dependent GVL effect, we demonstrate herein that CD8 effectors arising following immediate transfer induce more GVL than do those arising following delayed transfer. Effector CD8 cells arising following delayed transfer are recruited with slower kinetics, show a reduced capacity to generate IFN-γ, and, for CD8 cells bearing a high-affinity TCR, impaired viability. Both the lack of inflammation and extrinsic regulation by host cell populations contribute to the reduced competence of effector CTL arising in MC.

Animals were used under protocols approved by local institutional research committees and in accordance with National Institutes of Health or U.K. Home Office guidelines. Female BALB/cJ (H-2d), B10.A (H-2a), C57BL/6 (H-2b), B6 × DBA/2 (BDF1, H-2b/d), and B6.SJL (CD45.1) mice were purchased from the Frederick Cancer Research Facility (Frederick, MD). 2C TCR transgenic mice (H-2b on C57BL/6 background) (16) were kindly provided by Dr. Dennis Loh and bred at our facility. Donor mice were aged 6–13 wk and recipient mice were aged 11–12 wk at the time of transplantation.

Mixed or fully allogeneic hematopoietic chimeras were established as previously described (5, 17) by lethal irradiation followed by reconstitution with a mixture of T cell-depleted (TCD) donor (15 × 106) and recipient-type (5 × 106) bone marrow (BM) cells or TCD donor BM alone. Recipient mice were lethally irradiated (BDF1 at 11 Gy, B6 at 10.25 Gy, BALB/c at 8 Gy, 137Cs source, 0.8 Gy/min or x-ray irradiation, 0.55 Gy/min using same total dose but split in to two fractions, separated by 48 h) and TCD BM cells injected i.v. 4 h later.

Splenocyte (SC) suspensions were prepared as described (18), (19). Where required, naive donor 2C CD8 cells were isolated (purity >93%) by immunomagnetic selection using anti-CD8 microbeads (Miltenyi Biotec). In PCR experiments, transgenic CD8+1B2+ cells were isolated post-transfer by immunomagnetic negative selection of untouched T cells and positive selection using IgG1 1B2 clonotypic Ab and anti-mouse IgG1 microbeads (>95% purity). In experiments that involved transfer of T or non-T cell populations derived from established MC to secondary recipients, primary DLI recipients were sacrificed on day 12 and splenocytes harvested. T cells were purified by negative immunomagnetic selection to remove non-T cells (purity >93%, pan T isolation kit, Miltenyi Biotec) and the positive fraction used as a non-T cell population (90% purity). To select DLI-derived or host T cells, untouched CD3 cells were then incubated with biotinylated anti-CD45.1 Ab followed by positive selection with anti-biotin-conjugated microbeads (Miltenyi Biotec). CD45.1+CD3+ T cells constituted >97% of the final population. The negative fraction was used as the CD45.1CD3+ population.

The following mAbs were used: anti-CD4-PE (RM4-5), anti-CD8β-PE or -allophycocyanin (H35-17.2), anti-CD25-FITC (7D4), CD44-FITC (IM7), anti-CD45.2-FITC (104), anti-CD45.1-FITC or -biotin (A20), anti-IFN-γ-PE or -allophycocyanin (XMG1.2), anti-H2-Dd-FITC (34-2-12), anti-CD62L-FITC or -PE (MEL-14), anti-B220-PE (RA3-6B2), and the appropriate isotype controls (BD Pharmingen). Intracellular Foxp3 staining was performed using anti-Foxp3 (FJK-16s, eBioscience). Annexin VFITC (BD Pharmingen) was used to detect early apoptosis in propidium iodide-negative (live) cells. CD8 cells bearing the 2C TCR were identified via the clonotype-specific mAb 1B2 (20) and anti-mouse IgG1-allophycocyanin (X56, BD Pharmingen). Intracellular staining of cytokines was performed after overnight ex vivo stimulation with irradiated host or donor SC mixed at a 1:1 ratio in 96-well plates or following brief ex vivo stimulation with PMA/ionomycin. Flow cytometry was performed using a FACSCalibur flow cytometer (BD Biosciences).

Primers/conditions for quantitative RT-PCR and the analysis have been described previously (21) or are described in supplemental Table S1.6

Serum IFN-γ levels were measured in sera using a cytometric bead array kit (BD Biosciences) according to the kit instructions.

B cells were isolated from SC by immunomagnetic separation. Target cells were then differentially labeled with either low-dose (0.5 μM) or high-dose (5 μM) CFSE, mixed at a 1:1 ratio, and injected i.v. at a total of 10 × 106 cells/recipient. One to 16 h later cells recovered from the spleen were analyzed by flow cytometry for the presence of the two populations of CFSE-labeled cells and specific cytotoxicity was calculated as previously described (22).

EL4 cells were obtained from American Type Culture Collection and cultured as previously described (23).

Survival data were analyzed using the log rank test. Otherwise, statistical analyses were performed using Student’s t test (two-tailed).

To determine how the host environment might influence GVL, we first compared the degree of tumor protection following transfer of delayed DLI to MC vs early DLI to TBI-allo recipients. In these experiments, established B10.A + B6 → B6 MC received the same SC dose as a cohort of TBI-allo B6 mice also receiving a mixture of B6 and B10.A TCD BM. This assay measures a CD8 cell-mediated GVL effect against EL4, a class II-negative T cell lymphoma of host origin (5). Since early SC transfer to TBI-allo mice is associated with lethal GVHD, we performed preliminary titration experiments to determine the dose of donor B10.A SC that did not lead to prohibitive rates of GVHD in this setting (3 × 106 SC, not depicted). Having identified this limiting number of SC, we next transferred 3 × 106 B10.A SC on day 0 to groups of TBI-allo mice and to established MC followed by EL4 challenge via i.v. injection 1 wk later. Importantly, the 3 × 106 SC dose used in this experiment was 1 log lower than the SC dose (3 × 107 SC) required to induce GVL in MC (Ref. 5, 6, 7 and unpublished data). TBI-allo mice receiving EL4 and TCD BM alone on day 0 showed rapid mortality (Fig. 1, A and C median survival time (MST) 30 days), but mice receiving low-dose DLI and EL4 showed prolonged survival, even though two of nine mice died of GVHD as defined by lack of signs of tumor at autopsy (5) (MST of >90 days, p < 0.05). In contrast to TBI-allo mice, established MC (Fig. 1,B) receiving low-dose DLI with EL4 showed no significant improvement in survival over recipients of EL4 alone (MST of 28 days in both groups, Fig. 1, B and C). There were no differences in either the kinetics of EL4 tumor growth (see EL4/no DLI groups in Fig. 1, AC, p = NS) or its distribution (predominantly spleen, BM, liver) in MC or TBI-allo mice (not depicted). Thus, when using limiting numbers of SC, the GVL effect is greater upon transfer to TBI-allo mice compared with MC.

FIGURE 1.

Greater GVL effects following transfer of low numbers of donor splenocytes to TBI-allo recipients compared with established MC. A group of lethally irradiated B6 mice were reconstituted with a mixture of TCD B10.A plus TCD B6 BM cells and received 3 × 106 B10.A SC on day 56. On the same day, a group of TBI-allo B6 mice received mixed TCD B10.A and B6 BM plus the same number of B10.A SC. One week later 1 × 103 EL4 cells were injected to both groups. A, TBI-allo recipients: no DLI (▾, n = 4), DLI (♦, n = 7), EL4 (•, n = 5), or both (○, n = 9). Compared with tumor controls, mice receiving DLI + EL4 showed an improved survival rate (p < 0.05). B, Established MC: EL4 (□, n = 3) or DLI + EL4 (▪, n = 8). C, Tumor-free survival in TBI-allo mice vs established MC (p < 0.05). Group symbols as for A and B.

FIGURE 1.

Greater GVL effects following transfer of low numbers of donor splenocytes to TBI-allo recipients compared with established MC. A group of lethally irradiated B6 mice were reconstituted with a mixture of TCD B10.A plus TCD B6 BM cells and received 3 × 106 B10.A SC on day 56. On the same day, a group of TBI-allo B6 mice received mixed TCD B10.A and B6 BM plus the same number of B10.A SC. One week later 1 × 103 EL4 cells were injected to both groups. A, TBI-allo recipients: no DLI (▾, n = 4), DLI (♦, n = 7), EL4 (•, n = 5), or both (○, n = 9). Compared with tumor controls, mice receiving DLI + EL4 showed an improved survival rate (p < 0.05). B, Established MC: EL4 (□, n = 3) or DLI + EL4 (▪, n = 8). C, Tumor-free survival in TBI-allo mice vs established MC (p < 0.05). Group symbols as for A and B.

Close modal

To identify potential mechanisms underlying this marked difference between GVL activity in established MC and freshly irradiated recipients, we employed a model system that allowed us to track the fate of adoptively transferred T cells. Thus, 1 × 106 anti-host 2C CD8 cells bearing a transgenic TCR (with high affinity for host Ld Ag, identified using the clonotypic marker 1B2) were co-transferred together with 10 × 106 B6 CD45.1 SC to BALB/c TBI-allo recipients or established B6 + BALB/c → BALB/c MC. TBI-allo recipients were also given TCD B6 CD45.2 BM to prevent death from aplasia. Again, we employed numbers of T cells lower than previously used in MC to induce GVL (5, 6, 7) (in this case, 3-fold lower) to enable a comparison with TBI-allo recipients. This number of cells is sufficient to partially eradicate host hematopoietic elements in MC without inducing GVHD (5, 8).

As we have previously reported, 2C CD8 cells underwent rapid proliferation in both contexts, but with delayed accumulation within the spleen of MC and slower acquisition of a CD44highCD62Llow phenotype (8). To examine the levels of initial activation, we examined expression of the early activation marker, CD25 (Fig. 2,A). We observed almost no up-regulation of CD25 upon 2C CD8 cells following delayed DLI to MC at any time during the immune response, whereas >50% of cells expressed this receptor at day 3 following transfer to TBI-allo recipients (Fig. 2, A and B), indicating that the host environment influences early events in donor CD8 activation.

FIGURE 2.

Activation and effector differentiation of donor CD8 cells in TBI-allo recipients and MC. 2C CD8 cells (1 × 106) and 1 × 107 B6 CD45.1 SC were transferred to TBI-allo BALB/c or TBI-syn B6 recipients or B6 + BALB/c → BALB/c MC. A, Representative histograms of CD25 expression upon gated CD8+1B2+ cells on day 3. B, Graph showing CD25 expression at timed intervals post-transfer (n = 3/time point, ∗∗∗, p < 0.001 MC vs TBI-allo). Data from one of three independent experiments with similar results are shown. C, Serum levels of IFN-γ were measured at timed intervals posttransfer (n = 3/group/time point); ud indicates undetectable. Not done (nd) in TBI-allo cohort on day 10 or 12 as mice had succumbed to GVHD. D, Levels of IFN-γ and granzyme B mRNA were measured by quantitative RT-PCR in sorted transgenic T cells from spleen on days 6 and 12 following DLI to MCs (▪, n = 3–5) and TBI-allo recipients (○, with each point representing three pooled mice) in relation to expression levels in pooled naive 2C CD8 cells (shown as dotted line at level 1.0). One of two independent experiments with similar results is shown. E, 1 × 107 B6 CD45.1 SC (with or without TCD B6 CD45.2 BM) were transferred to TBI-allo BALB/c recipients or B6 + BALB/c → BALB/c MC. Day 6 IFN-γ synthesis in transferred CD8 T cells after overnight stimulation with irradiated BALB/c SC (n = 4/group, % DLI-derived CD8 cells IFN-γ+ 18.7 ± 3.2% in TBI-allo vs 0.2 ± 0.1% in MC, p = 0.001; absolute numbers IFN-γ+ 3.2 ± 0.1 × 106 in TBI-allo vs 0.03 ± 0.01 × 106 in MC, p = 0.02). F, In vivo cytotoxicity of CFSE-labeled BALB/c B cells within spleen on day 6 post-transfer (n = 3–5/group).

FIGURE 2.

Activation and effector differentiation of donor CD8 cells in TBI-allo recipients and MC. 2C CD8 cells (1 × 106) and 1 × 107 B6 CD45.1 SC were transferred to TBI-allo BALB/c or TBI-syn B6 recipients or B6 + BALB/c → BALB/c MC. A, Representative histograms of CD25 expression upon gated CD8+1B2+ cells on day 3. B, Graph showing CD25 expression at timed intervals post-transfer (n = 3/time point, ∗∗∗, p < 0.001 MC vs TBI-allo). Data from one of three independent experiments with similar results are shown. C, Serum levels of IFN-γ were measured at timed intervals posttransfer (n = 3/group/time point); ud indicates undetectable. Not done (nd) in TBI-allo cohort on day 10 or 12 as mice had succumbed to GVHD. D, Levels of IFN-γ and granzyme B mRNA were measured by quantitative RT-PCR in sorted transgenic T cells from spleen on days 6 and 12 following DLI to MCs (▪, n = 3–5) and TBI-allo recipients (○, with each point representing three pooled mice) in relation to expression levels in pooled naive 2C CD8 cells (shown as dotted line at level 1.0). One of two independent experiments with similar results is shown. E, 1 × 107 B6 CD45.1 SC (with or without TCD B6 CD45.2 BM) were transferred to TBI-allo BALB/c recipients or B6 + BALB/c → BALB/c MC. Day 6 IFN-γ synthesis in transferred CD8 T cells after overnight stimulation with irradiated BALB/c SC (n = 4/group, % DLI-derived CD8 cells IFN-γ+ 18.7 ± 3.2% in TBI-allo vs 0.2 ± 0.1% in MC, p = 0.001; absolute numbers IFN-γ+ 3.2 ± 0.1 × 106 in TBI-allo vs 0.03 ± 0.01 × 106 in MC, p = 0.02). F, In vivo cytotoxicity of CFSE-labeled BALB/c B cells within spleen on day 6 post-transfer (n = 3–5/group).

Close modal

Since we have previously reported that CD8 cell-derived IFN-γ is required for maximal GVL activity against EL4 tumor and for the eradication of host hematopoietic cells (24), we next measured the serum levels of this cytokine at timed intervals following T cell transfer. As shown in Fig. 2,C, IFN-γ levels were markedly elevated at days 3 and 6 in TBI-allo mice, but they were undetectable or very low in MC. Thereafter, we observed significant increases in IFN-γ levels in MC but at levels >5-fold lower than the peak observed in TBI-allo recipients. In parallel, we purified 2C transgenic T cells at various intervals from the spleens of recipient mice following their adoptive transfer, extracted RNA, and performed quantitative RT-PCR for a number of effector molecules whose expression was compared with that in naive 2C CD8 cells (Fig. 2,D and supplemental Table S2). In these experiments, cells were taken from mouse spleens on days 6 and 12 following adoptive transfer. The profiles generated revealed significant differences in the transcription of IFN-γ (Fig. 2,D). In two independent experiments, IFN-γ mRNA was up-regulated in 2C CD8 cells populations derived from both sets of mice, but there was a trend to lower peak expression in CTL derived from MC, and levels thereafter were not sustained to the same extent as in TBI-allo recipients (Fig. 2,D). In parallel, we examined the capacity of donor CD8 cells to generate IFN-γ following stimulation via their TCR with irradiated host BALB/c SC. Since such stimulation led to down-regulation of the 2C TCR making identification of such cells difficult, we instead examined IFN-γ synthesis by polyclonal DLI-derived CD8 cells. This population underwent similar blast formation and proliferation as CD8 cells bearing the transgenic TCR (not depicted). By day 6 there was a dramatic difference in the frequencies of transferred CD8 cells that expressed IFN-γ, with markedly increased IFN-γ production in cells derived from TBI-allo recipients (Fig. 2 E). This difference was not evident, however, when SC derived from MC and TBI-allo recipients were stimulated ex vivo with PMA and ionomycin (supplemental Fig. S1), suggesting that it was TCR proximal signaling events that distinguished the two populations at this time point.

Although we observed marked and sustained up-regulation of granzyme B mRNA in 2C cells in both cohorts of mice, the degree of up-regulation was clearly greater in the TBI-allo mice (Fig. 2 D). In contrast, no consistent differences were observed in the transcription of other cytokines, including IL-2, TNF-α, IL-4, or IL-10 mRNA (supplemental Table S2). FasL was not up-regulated in this system, and peak perforin mRNA expression was similar in both groups (supplemental Table S2). These differences were not the result of lymphopenia-induced proliferation, since the transcription of these molecules in 2C CD8 cells transferred to freshly irradiated syngeneic (TBI-syn) mice was similar to that of naive cells (not depicted).

We also examined the cytotoxic functions of donor T cells in vivo by evaluating the clearance of injected CFSEhigh host B cell targets (compared with co-injected CFSElow donor B cells). Host B cell eradication occurred rapidly (within 3 h, not depicted) and was virtually complete in TBI-allo mice on day 6, whereas it was absent at this early time point in MC even when analysis was delayed for 16 h (Fig. 2 F). Note, however, that host hematopoietic elements are partially eradicated in BALB/c MC given this intermediate dose of SC, although the kinetics of the response are over 14–28 days (8).

Taken together, these data show defects in the initial activation, effector differentiation, and cytotoxicity of anti-host CD8 T cells upon transfer to MC compared with freshly irradiated mice.

Since full GVL activity might also depend upon the survival of anti-host CD8 cells, we next examined the viability of 2C CD8 cells at timed intervals following their adoptive transfer and measured the degree of early apoptosis by annexin V (AV) staining (Fig. 3). No significant early apoptosis of 2C CD8 cells was observed in TBI-syn control mice. As previously reported (25), the contraction phase of the 2C CD8+ T cell response in TBI-allo recipients is associated with significant, early apoptosis (Fig. 3). By day 10, however, the proportion of cells undergoing apoptosis in this group had fallen to baseline. In sharp contrast, high rates of early apoptosis were observed at the initiation of the response and then were sustained until at least day 18 in MCs (Fig. 3), well beyond the peak of expansion at days 10–12 (Fig. 4). We observed similar findings following transfer of 2C cells and B6 SC to B6 + BDF1 → BDF1 chimeras (not depicted). These differences in viability were not observed for polyclonal CD8 T cells, however (not depicted).

FIGURE 3.

Reduced viability of 2C CD8 cells developing in MC compared with TBI-allo recipients. A and B, 1 × 106 2C CD8 cells and 1 × 107 B6 CD45.1 SC (with or without TCD B6 CD45.2 BM) were transferred to B6 + BALB/c → BALB/c MC (▪) or TBI-allo BALB/c (○) or TBI-syn B6 (•) recipients. At timed intervals, percentage annexin V (AV) binding in PI-negative CD8+1B2+ cells in recipient spleen was determined. Data shown are the proportions of CD8+1B2+ cells that were AV+ (A) or same data as absolute numbers (B). Note insufficient surviving mice in TBI-allo cohort from day 14 onward. Pooled data from three independent experiments are shown (∗, p < 0.05 and ∗∗∗, p < 0.001).

FIGURE 3.

Reduced viability of 2C CD8 cells developing in MC compared with TBI-allo recipients. A and B, 1 × 106 2C CD8 cells and 1 × 107 B6 CD45.1 SC (with or without TCD B6 CD45.2 BM) were transferred to B6 + BALB/c → BALB/c MC (▪) or TBI-allo BALB/c (○) or TBI-syn B6 (•) recipients. At timed intervals, percentage annexin V (AV) binding in PI-negative CD8+1B2+ cells in recipient spleen was determined. Data shown are the proportions of CD8+1B2+ cells that were AV+ (A) or same data as absolute numbers (B). Note insufficient surviving mice in TBI-allo cohort from day 14 onward. Pooled data from three independent experiments are shown (∗, p < 0.05 and ∗∗∗, p < 0.001).

Close modal
FIGURE 4.

Delayed and reduced BM recruitment of donor T cells generated in MC compared with TBI-allo recipients. 2C CD8 cells (1 × 106) and 1 × 107 B6 CD45.1 SC were transferred to TBI-allo BALB/c recipients or B6 + BALB/c → BALB/c MC. Irradiated recipients also received TCD B6 CD45.2 BM. At timed intervals posttransfer, the absolute numbers of transferred polyclonal CD8 and 2C CD8 T cells within the spleen, lymph node, and BM of MC (▪) and TBI-allo (○) recipient mice were determined (∗, p < 0.05; ∗∗, p < 0.01; ∗∗∗, p < 0.001; n = 3/group). Data from one of three independent experiments are shown.

FIGURE 4.

Delayed and reduced BM recruitment of donor T cells generated in MC compared with TBI-allo recipients. 2C CD8 cells (1 × 106) and 1 × 107 B6 CD45.1 SC were transferred to TBI-allo BALB/c recipients or B6 + BALB/c → BALB/c MC. Irradiated recipients also received TCD B6 CD45.2 BM. At timed intervals posttransfer, the absolute numbers of transferred polyclonal CD8 and 2C CD8 T cells within the spleen, lymph node, and BM of MC (▪) and TBI-allo (○) recipient mice were determined (∗, p < 0.05; ∗∗, p < 0.01; ∗∗∗, p < 0.001; n = 3/group). Data from one of three independent experiments are shown.

Close modal

Eradication of tumors also requires that CTL access sites of tumor development and metastasis. In this context, the pattern of recruitment of CTL to the secondary lymphoid organs and BM is likely to be crucial in eradication of many hematopoietic tumors. Using the same model system described in Fig. 2, we therefore examined the kinetics of accumulation of 2C and polyclonal donor CD8 T cells within these sites following transfer of DLI to MC or TBI-allo mice (Fig. 4; similar data relating to 2C CD8 cells for spleen and lymph node have been published previously (8) and are shown for comparison). In the spleen, the numbers of polyclonal donor CD8 cells or Ag-specific 2C CD8 cells were similar to or exceeded those in TBI-allo recipients. In contrast, within the BM there was a substantial delay in donor CD8 cell accumulation such that at day 6 polyclonal and 2C CD8+ T cell numbers were nearly 2 logs lower than those in TBI-allo recipients. Peak absolute 2C CD8+ numbers in MC BM were ∼5-fold lower than those in TBI-allo mice. These differences could not be accounted for by lymphopenia-induced proliferation at the different sites since similarly low levels of accumulation of donor T cells were observed in the BM of TBI-syn mice as in MC (not depicted). Findings in the lymph nodes were intermediate between those in the spleen and BM.

Taken together, the data in Figs. 2–4 indicate that the host environment is critical to the initial activation, effector differentiation, viability, and distribution of GVH-reactive CD8 cells.

The above findings were indicative of per-cell deficiencies in the functions of GVH-reactive CD8 cells activated in MC compared with TBI-allo recipients, at least at early time points. However, more detailed comparisons of donor CD8 effector functions over the whole primary response were difficult in the B6 → BALB/c strain combination since TBI-allo BALB/c recipients had a very high initial GVHD-related mortality (>50% at day 6). We therefore evaluated IFN-γ generation and cytotoxic functions of polyclonal CD8 cells following transfer of 1 × 107 B6 CD45.1 SC to TBI-allo BDF1 recipients or established B6 + BDF1 → BDF1 MC, in which GVHD lethality is observed at >14 days in the TBI-allo group (supplemental Fig. S2). In this strain combination, we also observed lower frequencies and absolute numbers of IFN-γ+ CD8 cells in MC compared with TBI-allo recipients even at later time points (supplemental Fig. S2). In contrast to the B6 → BALB/c strain combination, cytotoxicity against host targets cells was consistently detectable in B6 + BDF1 → BDF1 MC at day 6, albeit at reduced levels compared with TBI-allo recipients (supplemental Fig. S2). At day 14, in vivo cytotoxicity against host target cells was lower in B6 + BDF1 → BDF1 MC than in TBI-allo recipients. Despite these relative functional deficits, the 1 × 107 SC dose was sufficient to eradicate substantial numbers of host hematopoietic elements in MC but with kinetics over 14–28 days (not depicted).

Robust GVL responses in MC require transfer of very large numbers of SC (∼3 × 107 SC) (Refs. 5, 7 and our unpublished data). Increases in GVL in this context might simply reflect a linear dose-response effect. However, increasing the alloreactive precursor frequency could also potentially increase the per-cell effector functions of donor CD8 cells, perhaps through bystander effects. We therefore sought to determine how increasing the numbers of T cells infused would influence the functional properties of donor CD8 cells. Thus, we evaluated both IFN-γ generation by DLI-derived CD8 cells and the degree of in vivo cytotoxicity following transfer of 3 × 107 B6 CD45.1 SC to B6 + BDF1 → BDF1 MC (Fig. 5). Transfer of this dose of SC was associated with a marked accumulation of donor CD8 T cells in recipient spleen (Fig. 5,A), and by day 28, host hematopoietic elements had been eradicated almost completely despite the absence of GVHD (Fig. 5,B and not depicted). Increasing the number of T cells transferred was associated with no increase in the frequency of IFN-γ+ cells within the CD45.1+CD8+ population, but there was a detectable increase in their absolute numbers (Fig. 5, C and D). While cytotoxicity against host B cell targets at day 7 remained detectable but low, we now observed consistently high levels of cytotoxicity at days 14 and 30 (Fig. 5, E and F).

FIGURE 5.

Increasing SC dose compensates for per-cell deficits in donor CD8 T cells activated in MC. B6 CD45.1 SC (3 × 107) were transferred to B6 + BDF1 → BDF1 MC. A, Absolute numbers of CD8+CD45.1+ cells in spleen (n = 3/group/time point). B, Representative dot plots pretransfer and 28 days after SC transfer showing proportion of host (H-2d+) B220+ cells (n = 3/group). C, Absolute numbers of CD8+CD45.1+IFN-γ+ cells in recipient spleens. D, Representative dot plots of IFN-γ staining in gated CD45.1+ SC. E and F, In vivo cytotoxicity of CFSE-labeled BDF1 B cells within spleen at timed intervals posttransfer (n = 3–4/group).

FIGURE 5.

Increasing SC dose compensates for per-cell deficits in donor CD8 T cells activated in MC. B6 CD45.1 SC (3 × 107) were transferred to B6 + BDF1 → BDF1 MC. A, Absolute numbers of CD8+CD45.1+ cells in spleen (n = 3/group/time point). B, Representative dot plots pretransfer and 28 days after SC transfer showing proportion of host (H-2d+) B220+ cells (n = 3/group). C, Absolute numbers of CD8+CD45.1+IFN-γ+ cells in recipient spleens. D, Representative dot plots of IFN-γ staining in gated CD45.1+ SC. E and F, In vivo cytotoxicity of CFSE-labeled BDF1 B cells within spleen at timed intervals posttransfer (n = 3–4/group).

Close modal

Taken together, these data show that the transfer of larger numbers of T cells to MC compensates for but does not overcome the reduced per-cell function of donor CD8 activated in MC. In the absence of inflammation, enhancement of anti-host cytotoxicity and increases in the absolute numbers of anti-host CD8+IFN-γ+ cells did not, however, lead to GVHD.

We have previously demonstrated that induction of inflammation via TLR ligation at the time of T cell transfer to established B6 + BDF1 → BDF1 MC generates conditions permissive for the induction of GVHD (8). To model the effects of TLR triggering upon CD8 effector functions, B6 + BDF1 → BDF1 MC were treated with PBS or a CpG oligodeoxynucleotide (ODN 1826) on day −1 before and days 3 and 6 following transfer of 3 × 107 B6 CD45.1 SC. As expected (8), CpG-treated mice developed clinical signs of GVHD (not depicted). The phenotype of CD45.1+ CD8 cells was evaluated within recipient spleens on day 7. CpG treatment was associated with similar accumulation of donor CD8 cells in both groups (Fig. 6,A), with no increase in the donor CD8 cell generation of IFN-γ on a per-cell basis or in absolute numbers (Fig. 6,B and not depicted). However, in vivo cytotoxicity against host B cell targets on day 7 was enhanced dramatically by CpG treatment (Fig. 6,C) and this was associated with accelerated differentiation of dual CD44highCD62Llow CD8 cells (Fig. 6 D). Collectively, these data indicate that inflammation induced by a single TLR agonist is sufficient to accelerate the development of competent effector CTL in MC, although such cells continue to lack the full capacity to generate IFN-γ.

FIGURE 6.

Enhanced effector functions of donor CTL upon systemic TLR ligation. B6 CD45.1 SC (3 × 107) were transferred to B6 + BDF1 → BDF1 MC that received either PBS or CpG (100 μg ODN 1820) by i.p. injection on days −1, 3, and 6. A, Absolute numbers of CD8+CD45.1+ cells in spleen at day 7 (n = 4/group, p = NS). B, Day 6 IFN-γ synthesis in transferred CD8 T cells after overnight stimulation with irradiated BDF1 SC (n = 4/group, p = NS). C, In vivo cytotoxicity of CFSE-labeled BDF1 B cells within spleen on day 7 posttransfer (n = 4/group). D, Representative dot plots showing dual staining for CD62L and CD44 in CD8+CD45.1+ cells on day 6 posttransfer. Percentage cells CD44highCD62Llow: 85 ± 1% in CpG-treated and 71 ± 2% in control recipients (n = 4, p = 0.001). Data are pooled from two independent experiments with similar results.

FIGURE 6.

Enhanced effector functions of donor CTL upon systemic TLR ligation. B6 CD45.1 SC (3 × 107) were transferred to B6 + BDF1 → BDF1 MC that received either PBS or CpG (100 μg ODN 1820) by i.p. injection on days −1, 3, and 6. A, Absolute numbers of CD8+CD45.1+ cells in spleen at day 7 (n = 4/group, p = NS). B, Day 6 IFN-γ synthesis in transferred CD8 T cells after overnight stimulation with irradiated BDF1 SC (n = 4/group, p = NS). C, In vivo cytotoxicity of CFSE-labeled BDF1 B cells within spleen on day 7 posttransfer (n = 4/group). D, Representative dot plots showing dual staining for CD62L and CD44 in CD8+CD45.1+ cells on day 6 posttransfer. Percentage cells CD44highCD62Llow: 85 ± 1% in CpG-treated and 71 ± 2% in control recipients (n = 4, p = 0.001). Data are pooled from two independent experiments with similar results.

Close modal

In the absence of inflammation, one potentially important T cell extrinsic factor that may affect the development of CD8 effector function in MC is the presence of regulatory cell populations within the host. Previous work has shown that the GVH response following delayed transfer of T cells to BMT recipients can be ameliorated by both T (26, 27, 28, 29, 30) and non-T (31) regulatory cells. While we observed lower numbers of host and DLI-derived CD4+Foxp3+ cells following irradiation or CpG-induced inflammation compared with MC (supplemental Fig. S4), we were unable to detect significant suppression in MC following DLI using in vitro proliferation assays where non-T cell or T cell (CD45.1+ or CD45.2) fractions derived from the spleen were coincubated with normal syngeneic CD3+ cells stimulated with BDF1 SC (not depicted). We therefore employed the alternative strategy of adoptively cotransferring various cellular fractions to secondary, irradiated allogeneic recipients and testing function using the development of GVHD as a readout. We have previously demonstrated that transfer of selected DLI-derived T cells from MC to secondary allogeneic recipients leads to GVHD, indicating that there is no absolute intrinsic failure of effector differentiation in this population (8). In a first experiment, we transferred 2.5 × 107 B6 CD45.1 SC to established B6 + BDF1 → BDF1 MC. On day 12, recipient mice were sacrificed and splenocytes harvested. On the basis of one recipient spleen to one secondary recipient, 2.7 × 107 nylon wool-passaged SC were transferred to secondary freshly irradiated BDF1 recipients together with TCD B6 BM. In parallel, CD45.1+ SC derived from the DLI were sorted and transferred on the basis of one recipient to one secondary recipient (7 × 106 lymphocytes) to a second cohort of lethally irradiated recipients. In both cohorts of recipients we confirmed similar frequencies of transferred CD45.1+ CD8 cells in peripheral blood 16 days following transfer (not depicted). After a lag period of ∼20 days, secondary recipients of nonfractionated cells lost >10% weight and developed clinical signs of mild GVHD, but then showed almost complete recovery by day 60 post-BMT (Fig. 7,A). In sharp contrast, after a similar lag period, secondary recipients of CD45.1+-selected lymphocytes developed severe lethal GVHD (MST for CD45.1+ cells of 55 days vs >100 days for nonfractionated cells, p < 0.001). In a second series of experiments, recipient MC were sacrificed 12 days following DLI, and CD45.1+ T cells, CD45.1 T cells, and non-T cell fractions were isolated. TCD B6 BM together with 4 × 106 CD45.1+ purified T cells alone or in combination with recipient-derived 4 × 106 CD45.1 T cells or 12 × 106 non-T cells (reflecting their approximate ratios in recipient spleens) were then transferred to secondary lethally irradiated recipients (Fig. 7,B). Cotransfer of recipient CD45.1 T cells and, to a lesser extent, non-T cells reduced the incidence of lethal GVHD (Fig. 7 B). However, surviving mice receiving these accessory T or non-T cell fractions all demonstrated signs of histological GVHD (not depicted).

FIGURE 7.

Host T and non-T cell populations in MC impede the full development of GVHD-inducing effector function of donor T cells. A, Established B6 + BDF1 → BDF1 MC received 3 × 107 B6 CD45.1+ SC. On day 12, recipient mice were sacrificed and SC harvested. On the basis of one recipient spleen to one secondary recipient, 2.7 × 107 nylon wool-passaged SC were transferred to secondary freshly irradiated BDF1 recipients (○, n = 6). In parallel, CD45.1+ cells derived from the nylon-wool passaged cells were sorted and transferred on the basis of one recipient spleen to one secondary recipient (7 × 106 cells) to a second cohort of lethally irradiated recipients (▪, n = 6). All recipients received B6 CD45.2 TCD BM including an additional cohort of mice receiving no SC (□, n = 6). Clinical GVHD scores (left) and survival (right) are shown (MST CD45.1+ cells at 55 days vs >100 days nonfractionated cells, p < 0.001). B, Established B6 + BDF1 → BDF1 MC received 2.5 × 107 B6 CD45.1+ SC. On day 12, recipient mice were sacrificed, SC harvested, and CD3+CD45.1+, CD3+CD45.1, and CD3 (non-T cell) populations were sorted. Figure shows survival following transfer of 4 × 106 CD45.1+ T cells + TCD B6 BM (▪, n = 5), 4 × 106 CD45.1+, and 4 × 106 CD45.1 T cells + TCD B6 BM (▴, n = 5), or 4 × 106 CD45.1+ and 12 × 106 non-T cells + TCD B6 BM (▵, n = 5) to freshly irradiated secondary recipients (survival comparison CD45.1+ T cells vs CD45.1+/CD45.1 T cells, p = 0.002; CD45.1+ T cells vs CD45.1+/non-T cells, p < 0.05).

FIGURE 7.

Host T and non-T cell populations in MC impede the full development of GVHD-inducing effector function of donor T cells. A, Established B6 + BDF1 → BDF1 MC received 3 × 107 B6 CD45.1+ SC. On day 12, recipient mice were sacrificed and SC harvested. On the basis of one recipient spleen to one secondary recipient, 2.7 × 107 nylon wool-passaged SC were transferred to secondary freshly irradiated BDF1 recipients (○, n = 6). In parallel, CD45.1+ cells derived from the nylon-wool passaged cells were sorted and transferred on the basis of one recipient spleen to one secondary recipient (7 × 106 cells) to a second cohort of lethally irradiated recipients (▪, n = 6). All recipients received B6 CD45.2 TCD BM including an additional cohort of mice receiving no SC (□, n = 6). Clinical GVHD scores (left) and survival (right) are shown (MST CD45.1+ cells at 55 days vs >100 days nonfractionated cells, p < 0.001). B, Established B6 + BDF1 → BDF1 MC received 2.5 × 107 B6 CD45.1+ SC. On day 12, recipient mice were sacrificed, SC harvested, and CD3+CD45.1+, CD3+CD45.1, and CD3 (non-T cell) populations were sorted. Figure shows survival following transfer of 4 × 106 CD45.1+ T cells + TCD B6 BM (▪, n = 5), 4 × 106 CD45.1+, and 4 × 106 CD45.1 T cells + TCD B6 BM (▴, n = 5), or 4 × 106 CD45.1+ and 12 × 106 non-T cells + TCD B6 BM (▵, n = 5) to freshly irradiated secondary recipients (survival comparison CD45.1+ T cells vs CD45.1+/CD45.1 T cells, p = 0.002; CD45.1+ T cells vs CD45.1+/non-T cells, p < 0.05).

Close modal

Taken together, these data are consistent with a model in which host T and non-T cell populations in MC impede the full acquisition of GVHD-inducing effector functions by donor CD8 cells.

We have previously demonstrated that the host environment is critical in regulating the development of GVHD and that tissue inflammation is involved in the recruitment of activated T cells to nonlymphoid tissues (8). We have now addressed how the environment influences GVL and the differentiation of donor CD8 cells. Compared with the immediate transfer to TBI-allo recipients, delayed transfer of donor T cells to MC is associated with marked clonal expansion of GVH-reactive CD8 cells but delayed/incomplete effector differentiation and reduced cytotoxic functions. The absence of inflammation and the integrity of host regulatory populations in MC are important in restricting the full development of effector function and the development of GVHD. However, the reduced GVL activity of donor T cells on a per-cell basis in MC may be compensated for by the ability to transfer greater numbers of T cells without inducing GVHD.

In full allogeneic chimeras, increasing the interval between initial conditioning and the transfer of donor T cells leads to a decline in detectable GVH reactivity and eventual extinction of GVL (3, 4, 5). An important distinction in our study is that, by considering established mixed rather than full allogeneic chimeras, we have examined how delayed T cell transfer influences GVL under conditions in which the continued presence of host APCs enables priming and marked clonal expansion of anti-host CTL (5, 8). Our previous findings that anti-host T cells developing in MC can induce marked GVHD upon adoptive transfer to secondary allogeneic recipients or following induction of systemic inflammation indicate that such T cells do not have a complete, intrinsic deficit in function or survival (8). However, when limiting numbers of T cells were used in DLI (one tenth of the dose previously administered to MC to induce GVL) (5, 6, 7), we observed striking differences between MC and TBI-allo recipients in the magnitude of GVL. Although the host environment is also likely to influence the functions of both donor CD4 and CD8 cells, we have shown previously that GVL activity is CD4-independent in TBI-allo recipients (5). Thus, the present study suggests that differences in the effector differentiation of CD8 cells activated in MC and freshly irradiated hosts are at least partially responsible for the disparities observed in the magnitude of GVL.

Injury to intestinal and other epithelial barriers by TBI may allow entry of microbial products into the systemic circulation, triggering APC and T cell activation (11). Enhanced systemic levels of proinflammatory cytokines such as TNF-α, IL-1, and IL-6 (32) or costimulatory and cytokine expression by priming host dendritic cells (33) may accelerate the differentiation of CD8 effector cells in vivo (34). Massive egress of cells to inflamed, nonlymphoid tissues in TBI-allo hosts may further boost their lytic functions (35). Consistent with this possibility, artificial induction of inflammation with a TLR agonist was associated with accelerated CTL differentiation and greater killing of host target cells. Of note, however, was a failure to observe any significant increase in IFN-γ generation by donor CD8 cells activated in the presence of CpG, despite evidence of enhanced cytotoxicity. This contrasts with our finding that even very low doses of irradiation (2 Gy), which are expected to result in only very limited epithelial injury, are nevertheless sufficient to induce IFN-γhigh CTL (supplemental Fig. S4), suggesting that the effects of irradiation on CTL differentiation are pleiotropic and not dependent on the presence of inflammation alone.

Depletion or inhibition of regulatory cell populations (36, 37) by irradiation or inflammation could also be relevant to enhanced recruitment of polyfunctional CTL in TBI-allo recipients. Following delayed T cell transfer to allogeneic chimeras, both T cell (26, 27, 28, 29, 30) and non-T myeloid suppressor populations (31) are reported to restrict the development of GVHD. Although we were unable to detect regulatory activity using in vitro assays where individual cell populations were employed at ratios present in the spleens of MC, we did observe a failure of nylon wool-passaged SC preparations from DLI recipients to induce severe GVHD in secondary, irradiated allogeneic recipients and demonstrated that T cells present within MC before DLI were predominantly responsible for this effect. Although we have not defined the specific T cell subset(s) responsible for suppression in this model, these data are consistent with the finding that postthymic, donor-derived CD25+ T cells developing within allogeneic chimeras are important for restricting the development of GVHD (28, 29). Nevertheless, it is important to note that even in the face of this regulatory impedance, sufficient activation, proliferation, and effector CTL differentiation occur to permit significant lymphohematopoietic GVH reactivity and antitumor responses in MC.

The survival of functional anti-host CTL for a period of time sufficient to eradicate tumor is an essential prerequisite for a successful GVL response. In this study, anti-host CD8 cells bearing a high-affinity TCR (although not polyclonal CD8 cells) underwent sustained and high apoptotic frequencies in MC that were maintained well beyond the initiation of the contraction phase. Rapid loss of anti-host CD8 cells bearing high-affinity TCR in MC could be important in depleting the repertoire of T cells that can mediate effective GVL, although such cells are not completely deleted in this (our unpublished data) or similar models (6). The markedly reduced expression of CD25 (IL-2Rα) by anti-host CD8 cells in MC compared with TBI-allo recipients suggests the possibility that activated CD8 cells arising in the two contexts might have different capacities to respond to IL-2, a factor that may be of greater importance where there is intense competition for cytokines, as is likely to be the case in MC. The considerable expansion of 2C cells in MC despite their relative lack of expression of IL-2Rα is somewhat surprising since it has previously been shown that parental IL-2−/− SC fail to support expansion of 2C cells in a parent → F1 model (38). Our data would suggest therefore that IL-2 signaling (perhaps via the intermediate affinity receptor) is sufficient to promote primary expansion of 2C cells activated in MC. Priming in the absence of IL-2Rα expression might nevertheless lead to long-term dysfunction of such cells (39), a point that may be relevant to the failure to observe a GVL response to a secondary tumor challenge in MC given DLI (6).

The host environment regulates the initial distribution of donor CD8 cells within individual compartments of the lymphohematopoietic system. Thus, donor T cells show delayed kinetics of recruitment to the BM of MC compared with TBI-allo recipients. Such differences could be important in the tumor protection assay employed here, where initial tumor cell engraftment within the BM might escape surveillance by the early phase of alloreactive T cell response. Such findings might also help to explain the failure of DLI to induce responses in the acute leukemias (40), where rapid recruitment of CTL to the BM is likely to be important. During the early phase of the immune response in TBI-allo recipients, GVH-reactive CD8 cells recovered from the BM displayed a uniform CD44highCD62Llow phenotype, ruling out early site-specific selection for cells with a central memory like phenotype (41, 42) (not depicted). Although irradiation may induce “space” within the BM, it seems unlikely that this is the sole contributor to the different recruitment in the two contexts, since a saturable “niche” for recruitment of activated T cells to the BM was not demonstrated in a previous study (41). Indeed, escalation of the dose of donor SC transferred to MC leads to a proportionate increase in the numbers of donor CTL accumulating in the BM (not depicted). Although selectin, integrin, and chemokine requirements for central memory T cell trafficking to BM have begun to be identified (41), the determinants of recirculation of effector T cells to this site are not known in detail. We did not observe significant differences in the expression of CD44, VLA-4, or CXCR4 upon GVH-reactive CD8 cells in the MC or TBI-allo recipients (not depicted). An alternative explanation is that an inflammatory environment might induce changes within the BM vasculature that modulates access of activated T cells to, survival at, or egress from this site. For example, irradiation-induced injury to BM vasculature endothelium leads to SDF-1 up-regulation (43, 44) that could potentially promote T cell trafficking.

In summary, we have shown that the GVL response following delayed DLI to MC is associated with reduced functional competence and viability of effector CD8 cells. The feasibility of transferring larger numbers of T cells without inducing GVHD in this context may compensate for some of these per-cell deficiencies in effector function. These findings may therefore be relevant to protocols in human patients using delayed DLI following nonmyeloablative allogeneic hematopoietic stem cell transplantation, where the GVL effects may be suboptimal despite conversion to full donor chimerism (45). We have previously demonstrated that IFN-γ generation by donor CD8 cells is required for maximal GVL activity in freshly irradiated hosts and that this cytokine protects against the development of GVHD (24). Our data suggest that strategies to enhance GVL responses following delayed DLI to MC should center upon improving the capacity of donor CD8 cells to generate IFN-γ.

We thank Dr. Clare Bennett for helpful review of the manuscript. We also thank Orlando Moreno for expert oversight of animal care and Kelly Walsh for excellent assistance with the manuscript.

The authors have no financial conflicts of interest.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1

This work was supported by the Leukemia Research, U.K., and by National Institutes of Health Grants PO1 CA111519 and RO1 CA79989, and by a Senior Research Award from the Multiple Myeloma Foundation.

5

Abbreviations used in this paper: BMT, bone marrow transplantation; AV, annexin V; BM, bone marrow; DLI, donor leukocyte infusion; GVH, graft-vs-host; GVHD, graft-vs-host disease; GVL, graft-vs-leukemia; MC, mixed chimera; MST, median survival time; SC, splenocyte; TBI-allo, freshly irradiated allogeneic recipients; TBI-syn, freshly irradiated syngeneic recipients; TCD, T cell-depleted.

6

The online version of this article contains supplemental material.

1
Weiden, P. L., K. M. Sullivan, N. Flournoy, R. Storb, E. D. Thomas.
1981
. Antileukemic effect of chronic graft-versus-host disease: contribution to improved survival after allogeneic marrow transplantation.
N. Engl. J. Med.
304
:
1529
-1533.
2
Weiden, P. L., N. Flournoy, E. D. Thomas, R. Prentice, A. Fefer, C. D. Buckner, R. Storb.
1979
. Antileukemic effect of graft-versus-host disease in human recipients of allogeneic-marrow grafts.
N. Engl. J. Med.
300
:
1068
-1073.
3
Billiau, A. D., S. Fevery, O. Rutgeerts, W. Landuyt, M. Waer.
2002
. Crucial role of timing of donor lymphocyte infusion in generating dissociated graft-versus-host and graft-versus-leukemia responses in mice receiving allogeneic bone marrow transplants.
Blood
100
:
1894
-1902.
4
Xia, G., R. L. Truitt, B. D. Johnson.
2006
. Graft-versus-leukemia and graft-versus-host reactions after donor lymphocyte infusion are initiated by host-type antigen-presenting cells and regulated by regulatory T cells in early and long-term chimeras.
Biol. Blood Marrow Transplant.
12
:
397
-407.
5
Chakraverty, R., H. S. Eom, J. Sachs, J. Buchli, P. Cotter, R. Hsu, G. Zhao, M. Sykes.
2006
. Host MHC class II+ antigen-presenting cells and CD4 cells are required for CD8-mediated graft-versus-leukemia responses following delayed donor leukocyte infusions.
Blood
108
:
2106
-2113.
6
Mapara, M. Y., Y. M. Kim, J. Marx, M. Sykes.
2003
. Donor lymphocyte infusion-mediated graft-versus-leukemia effects in mixed chimeras established with a nonmyeloablative conditioning regimen: extinction of graft-versus-leukemia effects after conversion to full donor chimerism.
Transplantation
76
:
297
-305.
7
Mapara, M. Y., Y. M. Kim, S. P. Wang, R. Bronson, D. H. Sachs, M. Sykes.
2002
. Donor lymphocyte infusions mediate superior graft-versus-leukemia effects in mixed compared to fully allogeneic chimeras: a critical role for host antigen-presenting cells.
Blood
100
:
1903
-1909.
8
Chakraverty, R., D. Cote, J. Buchli, P. Cotter, R. Hsu, G. Zhao, T. Sachs, C. M. Pitsillides, R. Bronson, T. Means, et al
2006
. An inflammatory checkpoint regulates recruitment of graft-versus-host reactive T cells to peripheral tissues.
J. Exp. Med.
203
:
2021
-2031.
9
Cooke, K. R., A. Gerbitz, J. M. Crawford, T. Teshima, G. R. Hill, A. Tesolin, D. P. Rossignol, J. L. Ferrara.
2001
. LPS antagonism reduces graft-versus-host disease and preserves graft-versus-leukemia activity after experimental bone marrow transplantation.
J. Clin. Invest.
107
:
1581
-1589.
10
Guy-Grand, D., P. Vassalli.
1986
. Gut injury in mouse graft-versus-host reaction: study of its occurrence and mechanisms.
J. Clin. Invest.
77
:
1584
-1595.
11
Hill, G. R., J. M. Crawford, K. R. Cooke, Y. S. Brinson, L. Pan, J. L. Ferrara.
1997
. Total body irradiation and acute graft-versus-host disease: the role of gastrointestinal damage and inflammatory cytokines.
Blood
90
:
3204
-3213.
12
McCarthy, P. L., Jr, S. Abhyankar, S. Neben, G. Newman, C. Sieff, R. C. Thompson, S. J. Burakoff, J. L. Ferrara.
1991
. Inhibition of interleukin-1 by an interleukin-1 receptor antagonist prevents graft-versus-host disease.
Blood
78
:
1915
-1918.
13
Nestel, F. P., K. S. Price, T. A. Seemayer, W. S. Lapp.
1992
. Macrophage priming and lipopolysaccharide-triggered release of tumor necrosis factor α during graft-versus-host disease.
J. Exp. Med.
175
:
405
-413.
14
Piguet, P. F., G. E. Grau, B. Allet, P. Vassalli.
1987
. Tumor necrosis factor/cachectin is an effector of skin and gut lesions of the acute phase of graft-vs-host disease.
J. Exp. Med.
166
:
1280
-1289.
15
Johnson, B. D., R. L. Truitt.
1995
. Delayed infusion of immunocompetent donor cells after bone marrow transplantation breaks graft-host tolerance allows for persistent antileukemic reactivity without severe graft-versus-host disease.
Blood
85
:
3302
-3312.
16
Sha, W. C., C. A. Nelson, R. D. Newberry, D. M. Kranz, J. H. Russell, D. Y. Loh.
1988
. Positive and negative selection of an antigen receptor on T cells in transgenic mice.
Nature
336
:
73
-76.
17
Sykes, M., Z. Bukhari, D. H. Sachs.
1989
. Graft-versus-leukemia effect using mixed allogeneic bone marrow transplantation.
Bone Marrow Transplant.
4
:
465
-474.
18
Sykes, M., V. S. Abraham, M. W. Harty, D. A. Pearson.
1993
. IL-2 reduces graft-versus-host disease and preserves a graft-versus-leukemia effect by selectively inhibiting CD4+ T cell activity.
J. Immunol.
150
:
197
-205.
19
Sykes, M., M. W. Harty, G. L. Szot, D. A. Pearson.
1994
. Interleukin-2 inhibits graft-versus-host disease-promoting activity of CD4+ cells while preserving CD4- and CD8-mediated graft-versus-leukemia effects.
Blood
83
:
2560
-2569.
20
Kranz, D. M., S. Tonegawa, H. N. Eisen.
1984
. Attachment of an anti-receptor antibody to non-target cells renders them susceptible to lysis by a clone of cytotoxic T lymphocytes.
Proc. Natl. Acad. Sci. USA
81
:
7922
-7926.
21
Abdi, R., T. K. Means, A. D. Luster.
2003
. Chemokines in islet allograft rejection.
Diabetes Metab. Res. Rev.
19
:
186
-190.
22
Barber, D. L., E. J. Wherry, R. Ahmed.
2003
. Cutting edge: Rapid in vivo killing by memory CD8 T cells.
J. Immunol.
171
:
27
-31.
23
Yang, Y. G., J. J. Sergio, D. A. Pearson, G. L. Szot, A. Shimizu, M. Sykes.
1997
. Interleukin-12 preserves the graft-versus-leukemia effect of allogeneic CD8 T cells while inhibiting CD4-dependent graft-versus-host disease in mice.
Blood
90
:
4651
-4660.
24
Yang, Y. G., J. Qi, M. G. Wang, M. Sykes.
2002
. Donor-derived interferon gamma separates graft-versus-leukemia effects and graft-versus-host disease induced by donor CD8 T cells.
Blood
99
:
4207
-4215.
25
Dey, B., Y. G. Yang, F. Preffer, A. Shimizu, K. Swenson, D. Dombkowski, M. Sykes.
1999
. The fate of donor T-cell receptor transgenic T cells with known host antigen specificity in a graft-versus-host disease model.
Transplantation
68
:
141
-149.
26
Blazar, B. R., C. J. Lees, P. J. Martin, R. J. Noelle, B. Kwon, W. Murphy, P. A. Taylor.
2000
. Host T cells resist graft-versus-host disease mediated by donor leukocyte infusions.
J. Immunol.
165
:
4901
-4909.
27
Gonzalez, M., S. A. Quezada, B. R. Blazar, A. Panoskaltsis-Mortari, A. Y. Rudensky, R. J. Noelle.
2002
. The balance between donor T cell anergy and suppression versus lethal graft-versus-host disease is determined by host conditioning.
J. Immunol.
169
:
5581
-5589.
28
Johnson, B. D., E. E. Becker, J. L. LaBelle, R. L. Truitt.
1999
. Role of immunoregulatory donor T cells in suppression of graft-versus-host disease following donor leukocyte infusion therapy.
J. Immunol.
163
:
6479
-6487.
29
Johnson, B. D., M. C. Konkol, R. L. Truitt.
2002
. CD25+ immunoregulatory T-cells of donor origin suppress alloreactivity after BMT.
Biol. Blood Marrow Transplant.
8
:
525
-535.
30
Young, K. J., B. DuTemple, M. J. Phillips, L. Zhang.
2003
. Inhibition of graft-versus-host disease by double-negative regulatory T cells.
J. Immunol.
171
:
134
-141.
31
Billiau, A. D., S. Fevery, O. Rutgeerts, W. Landuyt, M. Waer.
2003
. Transient expansion of Mac1+Ly6-G+Ly6-C+ early myeloid cells with suppressor activity in spleens of murine radiation marrow chimeras: possible implications for the graft-versus-host and graft-versus-leukemia reactivity of donor lymphocyte infusions.
Blood
102
:
740
-748.
32
Xun, C. Q., J. S. Thompson, C. D. Jennings, S. A. Brown, M. B. Widmer.
1994
. Effect of total body irradiation, busulfan-cyclophosphamide, or cyclophosphamide conditioning on inflammatory cytokine release and development of acute and chronic graft-versus-host disease in H-2-incompatible transplanted SCID mice.
Blood
83
:
2360
-2367.
33
Zhang, Y., J. P. Louboutin, J. Zhu, A. J. Rivera, S. G. Emerson.
2002
. Preterminal host dendritic cells in irradiated mice prime CD8+ T cell-mediated acute graft-versus-host disease.
J. Clin. Invest.
109
:
1335
-1344.
34
Badovinac, V. P., K. A. Messingham, A. Jabbari, J. S. Haring, J. T. Harty.
2005
. Accelerated CD8+ T-cell memory and prime-boost response after dendritic-cell vaccination.
Nat. Med.
11
:
748
-756.
35
Marzo, A. L., H. Yagita, L. Lefrançois.
2007
. Cutting edge: Migration to nonlymphoid tissues results in functional conversion of central to effector memory CD8 T cells.
J. Immunol.
179
:
36
-40.
36
Caramalho, I., T. Lopes-Carvalho, D. Ostler, S. Zelenay, M. Haury, J. Demengeot.
2003
. Regulatory T cells selectively express Toll-like receptors and are activated by lipopolysaccharide.
J. Exp. Med.
197
:
403
-411.
37
Pasare, C., R. Medzhitov.
2003
. Toll pathway-dependent blockade of CD4+CD25+ T cell-mediated suppression by dendritic cells.
Science
299
:
1033
-1036.
38
Buhlmann, J. E., M. Gonzalez, B. Ginther, A. Panoskaltsis-Mortari, B. R. Blazar, D. L. Greiner, A. A. Rossini, R. Flavell, R. J. Noelle.
1999
. Cutting edge: Sustained expansion of CD8+ T cells requires CD154 expression by Th cells in acute graft versus host disease.
J. Immunol.
162
:
4373
-4376.
39
Williams, M. A., A. J. Tyznik, M. J. Bevan.
2006
. Interleukin-2 signals during priming are required for secondary expansion of CD8+ memory T cells.
Nature
441
:
890
-893.
40
Kolb, H. J., C. Schmid, A. J. Barrett, D. J. Schendel.
2004
. Graft-versus-leukemia reactions in allogeneic chimeras.
Blood
103
:
767
-776.
41
Mazo, I. B., M. Honczarenko, H. Leung, L. L. Cavanagh, R. Bonasio, W. Weninger, K. Engelke, L. Xia, R. P. McEver, P. A. Koni, L. E. Silberstein, U. H. von Andrian.
2005
. Bone marrow is a major reservoir and site of recruitment for central memory CD8+ T cells.
Immunity
22
:
259
-270.
42
Zeng, D., P. Hoffmann, F. Lan, P. Huie, J. Higgins, S. Strober.
2002
. Unique patterns of surface receptors, cytokine secretion, and immune functions distinguish T cells in the bone marrow from those in the periphery: impact on allogeneic bone marrow transplantation.
Blood
99
:
1449
-1457.
43
Bastianutto, C., A. Mian, J. Symes, J. Mocanu, N. Alajez, G. Sleep, W. Shi, A. Keating, M. Crump, M. Gospodarowicz, et al
2007
. Local radiotherapy induces homing of hematopoietic stem cells to the irradiated bone marrow.
Cancer Res.
67
:
10112
-10116.
44
Ponomaryov, T., A. Peled, I. Petit, R. S. Taichman, L. Habler, J. Sandbank, F. Arenzana-Seisdedos, A. Magerus, A. Caruz, N. Fujii, et al
2000
. Induction of the chemokine stromal-derived factor-1 following DNA damage improves human stem cell function.
J. Clin. Invest.
106
:
1331
-1339.
45
Peggs, K. S., K. Thomson, D. P. Hart, J. Geary, E. C. Morris, K. Yong, A. H. Goldstone, D. C. Linch, S. Mackinnon.
2004
. Dose-escalated donor lymphocyte infusions following reduced intensity transplantation: toxicity, chimerism, and disease responses.
Blood
103
:
1548
-1556.