Report

Efficient Transplantation via Antibody-Based Clearance of Hematopoietic Stem Cell Niches

See allHide authors and affiliations

Science  23 Nov 2007:
Vol. 318, Issue 5854, pp. 1296-1299
DOI: 10.1126/science.1149726

Abstract

Upon intravenous transplantation, hematopoietic stem cells (HSCs) can home to specialized niches, yet most HSCs fail to engraft unless recipients are subjected to toxic preconditioning. We provide evidence that, aside from immune barriers, donor HSC engraftment is restricted by occupancy of appropriate niches by host HSCs. Administration of ACK2, an antibody that blocks c-kit function, led to the transient removal of >98% of endogenous HSCs in immunodeficient mice. Subsequent transplantation of these mice with donor HSCs led to chimerism levels of up to 90%. Extrapolation of these methods to humans may enable mild but effective conditioning regimens for transplantation.

Allogeneic bone marrow transplantation (BMT) generally requires conditioning of the recipient through cytoreductive treatments to prevent immunological rejection of the graft. Because these regimens can be associated with serious side effects (1), such preparative treatments are often omitted for patients with diseases such as severe combined immunodeficiency (SCID), as these patients are incapable of rejecting donor grafts (2). Nonetheless, although large numbers of B and T lymphocytes are at least transiently generated, the levels of donor HSC engraftment are usually less than 1% after transplantation into unconditioned SCID recipients (36). Although several studies have concluded that transplanted HSCs can easily replace endogenous HSCs without conditioning (7, 8), earlier studies suggested that the availability of niches is a limiting factor to transplantation (9). Thus, we reasoned that donor HSC engraftment might be limited by the occupancy of appropriate niches by endogenous HSCs, and that the development of reagents that specifically remove host HSCs could lead to safer transplantation-based therapies for hematological and nonhematological disorders than those currently in use.

To address whether endogenous HSCs could be replaced by transplanted HSCs in a linear dose-dependent manner, we transplanted unconditioned Rag2–/– γc–/– mice (10, 11) with varying numbers of c-kit+ lineage Sca-1+ (KLS) CD34 CD150+ HSCs from mice expressing green fluorescent protein (GFP) driven by the β-actin promoter (1216). Peripheral blood granulocyte chimerism was measured at 16 weeks after transplant; this method has been shown to accurately reflect donor HSC chimerism in this system (3, 15). Donor granulocyte chimerism increased measurably in doses of 10 to 250 transplanted HSCs, but transplantation of more than 250 cells led to at most modest increases in chimerism (Fig. 1). Transplantation of 18,000 HSCs, representing ∼70% of the total number of HSCs in an adult mouse (17, 18), led to a mean chimerism of only 3% (Fig. 1, top panel). Similar results were obtained through transplantation of unfractionated bone marrow (12) (fig. S1). These data suggest that without conditioning, HSC engraftment is limited by the number of saturable niches that are empty at the time of transplant or become empty as the transplanted cells still survive.

Fig. 1.

Available HSC niches can be saturated with donor HSCs. Peripheral blood of transplanted unconditioned RAG2–/– γc–/– mice was analyzed 16 weeks after HSC transplantation for GFP+ granulocytes. The lower panel represents an expanded view of results from transplantation of 10 to 1000 HSCs. In the lower right panel, mice were cotransplanted with 1000 CD45.1 HSCs and 100,000 GFP+ KLS CD34+ cells. Mean values ± SEM are shown (n = 4 or 5 for each dose); **P <0.05 relative to the chimerism arising from the 10 HSC–transplanted group. The dashed line represents the theoretical HSC chimerism if engraftment were to increase linearly with transplanted cell dose from the observed chimerism at the 50 HSC–dose group.

To determine the specificity of these niches, we competitively transplanted unconditioned Rag2–/– γc–/– (CD45.2) mice with 1000 CD45.1 HSCs along with 100,000 GFP+ KLS CD34+ progenitor cells, which are the immediate progeny of HSCs (16, 19). No significant difference in donor HSC chimerism was observed in these mice relative to recipients that received 1000 HSCs alone (P = 0.77) (Fig. 1, lower right panel), which suggests that HSCs and their immediate progeny use distinct niches to maintain function.

To determine whether the specific elimination of host HSCs would allow for high levels of donor HSC engraftment, we compared a number of different candidate HSC-depleting monoclonal antibodies, including α-Sca1 (13), α-integrin α4 (20), and α-ESAM1 (21), but ultimately selected ACK2, an antibody known to recognize and antagonize c-kit (22), the receptor for stem cell factor (SCF) (23). We reasoned that if the ACK2 antibody were capable of depleting endogenous HSCs, residual antibody in the serum of mice would also deplete transplanted donor HSCs. To determine the kinetics of antibody clearance in vivo, we administered ACK2 intravenously to Rag2–/– γc–/– mice and tested the serum for the presence of antibody (12). ACK2 remained in the serum until 5 days after injection; however, no ACK2 was detected by 7 or 8 days after injection (Fig. 2A). All mice survived the ACK2 treatment with no obvious signs of distress (12) (table S1 and fig. S2). At this time point, a ∼99% decrease in the number of HSCs (Fig. 2B) was observed, and comparable levels of depletion were observed in B cell–deficient and wild-type mice (12) (fig. S3). To further confirm HSC depletion, we transplanted 200,000 bone marrow cells into lethally irradiated recipients 9 days after ACK2 treatment (12). Transplantation of bone marrow from ACK2-treated animals led to >90% lower engraftment relative to controls (Fig. 2C), thus confirming that ACK2 depletes functional HSCs from the bone marrow.

Fig. 2.

ACK2 treatment depletes HSCs in vivo. (A) ACK2 is cleared from serum of RAG2–/– γc–/– mice 7 days after injection. Serum of mice receiving 500 μg of ACK2 was analyzed every 2 days for ACK2 antibody by staining c-kit+ mast cells (31). (B) ACK2 administration leads to depletion of bone marrow HSCs. Numbers of KLS CD135 CD150+ HSCs were determined in femurs and tibia of ACK2-treated and control mice. Mean values ± SEM are shown (n = 3 for each time point); **P < 0.001. (C) ACK2 treatment, but not 2B8 treatment, depletes functional HSCs from bone marrow. We transplanted 200,000 unfractionated bone marrow cells from RAG2–/– γc–/– mice, treated with 500 μg of ACK2 or 2B8 9 days earlier, into wild-type irradiated recipients alongside 200,000 untreated competitor wild-type bone marrow cells. Mean values ± SEM are shown (n = 5 to 8); **P < 0.01. (D) ACK2 treatment does not directly cause HSC mobilization to the spleen. Entire splenocyte populations from mice treated with 500 μg of ACK2 9 days earlier were transplanted alongside 200,000 competitor bone marrow cells from wild-type mice. Mean values ± SEM are shown (n = 3 to 9); **P < 0.001. (E) ACK2 inhibits SCF-mediated HSC proliferation. HSCs were isolated from wild-type mice, plated at 10 cells per well cultured in the presence of SCF or TPO and ACK2 or 2B8. Proliferation was observed by light microscopy. **P < 0.05 as compared to ACK2-treated samples.

To determine the mobilizing effects of ACK2 treatment, we transplanted the entire splenocyte populations of ACK2-conditioned animals into lethally irradiated recipients (12). The donor chimerism resulting from ACK2-treated splenocytes was reduced by >95% relative to controls (Fig. 2D). These data indicate that ACK2 treatment does not induce HSC depletion from the bone marrow by mobilizing HSCs to the spleen. Moreover, we were unable to detect HSCs in the spleen, liver, or blood after ACK2 treatment by flow cytometry (fig. S4). Mobilization was only observed during the recovery phase, well after ACK2 had already cleared from the serum (fig. S5).

To determine the mechanism by which ACK2 depletes HSCs, we compared the effects of ACK2 treatment to that of 2B8, another c-kit monoclonal antibody of the same IgG2b isotype. 2B8 treatment did not decrease functional HSC numbers in vivo, as transplantation of bone marrow cells from mice treated with 2B8 resulted in levels of engraftment similar to those of controls (Fig. 2C). Next, we cultured purified HSCs in the presence of ACK2 and found that it completely inhibited SCF-dependent proliferation, but not thrombopoietin (TPO)–mediated proliferation (Fig. 2E) (12). These data, along with data showing that distinct c-kit–expressing progenitors are depleted at differential rates (12) (fig. S6), suggest that ACK2 does not deplete through Fc-mediated functions. Rather, consistent with studies using c-kit mutant mice (12, 24, 25), the complete inhibition of c-kit signaling by ACK2 (Fig. 2E) can deplete HSCs, whereas the partial inhibition by 2B8 cannot.

By 23 days after treatment, HSC cell surface profiles (fig. S5) and numbers (Fig. 2B) had returned to near normal levels. These data indicate that ACK2 causes a marked but transient depletion of host HSCs and results in a short window during which ACK2-treated animals might be receptive to donor HSC transplantation. To test whether the ablation of host HSCs could improve the efficiency of donor HSC engraftment, we conditioned RAG2–/– (CD45.1) mice with ACK2 and transplanted them with 5000 wild-type CD45.2 HSCs after serum clearance of ACK2. The mean donor granulocyte chimerism at 37 weeks after transplant was 16.1%, reflecting a factor of >10 increase over control recipients (Fig. 3A). These engrafted HSCs also gave rise to peripheral B and T cells (Fig. 3B).

Fig. 3.

ACK2 treatment enhances HSC engraftment. (A) ACK2 conditioning leads to higher donor myeloid chimerism. Donor granulocyte chimerism was measured after transplantation of 5000 HSCs in RAG2–/– mice conditioned with ACK2 7 days before transplant and compared to that of unconditioned mice. Mean values ± SEM are shown (n = 4); **P < 0.01. (B) HSC transplantation of ACK2-treated animals leads to lymphocyte reconstitution. We enumerated splenic donor-derived B and T cells from ACK2-treated and unconditioned RAG2–/– mice 39 weeks after transplantation with wild-type HSCs. Mean values ± SEM are shown (n = 3 to 5); **P < 0.01. (C) Granulocyte chimerism accurately measures bone marrow HSC chimerism. Peripheral blood granulocyte (Ter119 CD3 B220 Mac-1high side scatterhigh) chimerism at 37 weeks after transplantation was correlated with HSC (c-kit+ lineage Sca-1+ CD34 CD150+) chimerism in the bone marrow at 39 weeks after transplantation upon killing. Solid line illustrates linear regression, with 95% confidence interval shaded in gray. Dashed line represents theoretical values if donor granulocyte chimerism were identical to donor HSC chimerism. (D) Secondarily transplanted donor HSCs from ACK2-treated mice give rise to long-term multilineage engraftment. Peripheral blood chimerism of B cells (B), T cells (T), and granulocytes (G) is shown 16 weeks after secondary transplant for two independent experiments. Mean values ± SEM are shown (n = 7 or 8 in each experiment).

We also analyzed bone marrow HSC chimerism directly at this late time point to confirm the increase in engraftment, and found that it correlated well with peripheral blood granulocyte chimerism (Fig. 3C). Secondary transplants of donor HSCs re-isolated from primary recipients led to multilineage engraftment for at least 16 weeks after transplant (Fig. 3D), confirming that transplanted HSCs regain their normal cell surface phenotype and cell cycle status (fig. S7) by at least 7 to 9 months after transplant in ACK2-treated animals (12).

These data did not distinguish whether ACK2 treatment increased niche space or whether, as in unconditioned animals, only a small fraction of transplanted HSCs initially engrafted but then competitively expanded. If niche space had truly been freed, HSC chimerism would be expected to increase linearly with transplanted cell dose. To distinguish between these alternatives, we transplanted ACK2-conditioned RAG2–/– γc–/– (CD45.2) recipient mice with varying doses of CD45.1 HSCs. Donor engraftment increased linearly with transplanted HSC dose in two independent experiments at both early and late time points (Fig. 4A), consistent with the emptying of HSC niches by ACK2 treatment. Strikingly, donor chimerism values of up to 90% were achieved at the 35,000 HSC dose (Fig. 4B) and upon repetitive rounds of ACK2 treatment and transplantation of a total of 15,000 HSCs (fig.S8), which, upon extrapolation to humans, is a clinically obtainable number (12).

Fig. 4.

Donor chimerism increases with transplanted HSC cell number in ACK2-treated mice. (A) ACK2 treatment increases available HSC niche space. In two separate experiments, RAG2–/– γc–/– mice were treated with ACK2 and transplanted 9 days later with varying doses of HSCs (CD45.1). Donor granulocyte chimerism was measured as above 24 weeks after transplantation for the first experiment, and 4 weeks after transplantation for the second experiment. Mean values ± SEM are shown. (B) Flow cytometry profiles of mice transplanted with 35,000 HSCs. Chimerism of CD3 B220 Mac1high side scatterhigh peripheral blood granulocytes is shown. Numerical values represent percent donor chimerism.

Allogeneic BMT is used routinely for a number of clinical purposes, such as for the treatment of SCID (26, 27). Our results provide evidence that in the absence of conditioning, donor HSC engraftment is limited by the occupancy of appropriate niches by host HSCs. These data offer an explanation for the poor donor HSC engraftment observed in unconditioned SCID patients (4, 5), whichinturnmay be responsiblefor thelow levels of donor B lymphopoiesis and the finite duration of T cell production (6, 28).

We have shown that administration of ACK2 in vivo leads to the rapid but transient depletion of host HSCs, and that subsequent transplantation of highly purified HSCs leads to donor chimerism levels of up to 90%. When coupled with highly specific immunosuppressive depleting antibodies, shown tobe effective inbothmice (29) and humans (30) as transplantation conditioners, the use of HSC-specific depleting antibodies may be an attractive alternative to conventional methods of conditioning [which carry serious health risks (1)] and may thus increase the utility of allogeneic BMT for both hematological and nonhematological disorders.

Supporting Online Material

www.sciencemag.org/cgi/content/full/318/5854/1296/DC1

Materials and Methods

SOM Text

Figs. S1 to S7

Table S1

References

References and Notes

View Abstract

Navigate This Article