Preformed antibody, not primed T cells, is the initial and major barrier to bone marrow engraftment in allosensitized recipients

Abstract

Multiply-transfused individuals are at higher risk for BM rejection. We show that whereas allosensitization resulted in the priming of both cellular and humoral immunity, preformed antibody was the major barrier to engraftment. The generation of cross-reactive alloantibody led to rejection of BM of a different MHC-disparate strain. Imaging studies indicated that antibody-mediated rejection was very rapid (<3 hours) in primed recipients, while T-cell-mediated rejection in nonprimed mice took more than 6 days. Antibody-mediated BM rejection was not due to a defect in BM homing as rejection occurred despite direct intra-BM infusion of donor BM. Rejection was dependent upon host FcR+ cells. BM cells incubated with serum from primed mice were eliminated in nonprimed recipients, indicating that persistent exposure to high-titer antibody was not essential for rejection. High donor engraftment was achieved in a proportion of primed mice by mega-BM cell dose, in vivo T-cell depletion, and high-dose immunoglobulin infusion. The addition of splenectomy to this protocol only modestly added to the efficacy of this combination strategy. These data demonstrate both rapid alloantibody-mediated elimination of BM by host FcR+ cells and priming of host antidonor T cells and suggest a practical strategy to overcome engraftment barriers in primed individuals.

Figure 1

Allosensitization results in antibody-mediated donor BM rejection. (A,C) B6 mice were lethally irradiated on day −1 (8.0 Gy) and infused with 20 × 10⁶ (A) or 40 × 10⁶ (C) BALB/c BM on day 0. Mice were primed at various times prior to BMT by the intraperitoneal administration of 20 × 10⁶ BALB/c splenocytes. One group received anti-CD4, anti-CD8, and anti-NK mAbs around the time of transplantation (A). Survival is shown. (A) n = 16 to 26/group; data from 2 to 3 separate experiments with similar results were pooled. (C) n = 5/group. (B,D) Serum was collected from BALB/c-primed B6 mice after the indicated interval, diluted, and incubated with BALB/c thymocytes. Cells were washed and incubated with FITC-conjugated goat anti–mouse Ig Ab and analyzed by flow cytometry. Overlay histograms indicate alloantibody in diluted serum from 5 different mice binding to BALB/c thymocytes. Negative control of FITC-conjugate binding to thymocytes in absence of serum is indicated by the thin dotted line. Average mean fluorescent intensity (MFI) is listed for panel D. (D) *P < .05 versus primed day −28; n = 5. Each bold line represents a different mouse.

Figure 2

Priming to one alloantigen can result in the elimination of third-party BM indicating nonspecificity in the priming response. (A) B6 mice were primed against BALB/c or B10.BR (BR) by the intraperitoneal administration of 20 × 10⁶ splenocytes, lethally irradiated on day −1 (8.0 Gy), and infused with BALB/c or BR BM (20 × 10⁶) on day 0. Survival is shown. n = 8/group. (B) Serum was pooled from BALB/c- or B10.BR-primed B6 mice (n = 5) and incubated with BALB/c or B10.BR BM. Histograms illustrate binding of serum antibody to BM of priming strain (top panels) and third-party strain (bottom panels). Percentage positive binding is given. Negative control is shown by thin dashed line.

Figure 3

The ex vivo incubation of donor BALB/c BM with serum from BALB/c-primed B6 mice results in the destruction of the antibody-coated donor BM cells in a nonprimed B6 recipient. (A) Serum was collected from naive or BALB/c-primed mice, diluted 1:20, and incubated with BALB/c or BR BM. BM was thoroughly washed. An aliquot was incubated with FITC-conjugated goat anti–mouse Ig Ab and analyzed by flow cytometry. Histograms indicate binding of serum alloantibody to BALB/c or BR BM. Negative control of FITC-conjugate binding to BM in absence of serum is indicated by the thin dotted line. (B) Nonprimed B6 mice were lethally irradiated (8.0 Gy) and infused with BALB/c or BR BM cells (10 × 10⁶) that had been incubated with serum from naive or BALB/c-primed B6 mice as shown in panel A. Survival is shown. n = 8 to 10/group; BALB/c BM incubated with naive versus BALB/c-primed serum, P = .003. Results were reproduced in a second experiment.

Figure 4

Antibody-mediated rejection of donor BM in primed mice is far more rapid than T-cell–mediated rejection in naive mice. Shown are images of BALB/c mice irradiated with 3.5 Gy on day −1 and infused with 20 × 10⁶ B6 GFP⁺ T-cell–depleted BM cells on day 0. (A-G) Left panels (rejection) indicate untreated mice that ultimately reject their grafts in a T-cell–dependent fashion. Middle panels (engraftment) indicate mice that received anti-CD4 and anti-CD8 mAbs in vivo around the time of transplantation to ensure long-term donor engraftment. Right panels (primed rejection) indicate mice primed with B6 splenocytes on day −28 that reject their donor BM grafts via preformed alloantibody. (H-I) Nonprimed and primed mice as indicated were imaged 3 hours after BMT. Representative images from 1 of 3 mice are shown. See “Materials and methods” for imaging details.

Figure 5

Antibody-mediated BM rejection in the primed recipient is dependent on host FcR⁺ cells. B6 and B6.Fcϵr1γ^−/− mice were primed against BALB/c on day −28, lethally irradiated (8.0 Gy) on day −1, and infused with 20 × 10⁶ BALB/c T-cell–depleted BM cells on day 0. A cohort received anti-CD4 and anti-CD8 mAbs around the time of transplantation to ensure depletion of host T cells. (A) Overlay histograms illustrate equivalent serum alloantibody levels at time of BMT in primed Fcϵr1γ^−/− as wild-type mice in a thymocyte-binding assay. Serum from nonprimed mice (histograms on the left) illustrates low-level binding. Histograms on the right illustrate high degree of thymocyte binding by serum from primed mice. n = 5/group. Negative control (no serum) is indicated by the thin dotted line. (B) Survival is shown. n = 10/group; P < .001 for primed B6 versus primed B6.Fcϵr1γ^−/−. With the exception of the group of primed B6.FcϵrIγ^−/− mice that received anti-CD4 and anti-CD8 mAbs, data were reproduced in a second experiment. Each bold line represents a different mouse.