Search

We adapted organotypic skin cultures to the dog as a model for skin graft-versus-host reaction (GVHR) to explore the relative roles of T cells and cytokines. To produce GVHR, activated lymphocytes from bulk mixed leukocyte cultures (bMLC) from 2 dog leukocyte antigen-unrelated dogs were injected into organotypic skin cultures. Additionally, effects of separated CD4+ and CD8+ activated lymphocytes as well as cytokine-containing (TNF alpha and IFN gamma) supernatants from bMLC were studied. Noninjected cultures as well as cultures injected with autologous (cultured and uncultured) lymphocytes and allogeneic uncultured lymphocytes served as controls. The unseparated bMLC-activated cell populations induced histopathological changes similar to in vivo skin GVHR along with very prominent class II antigen expression on keratinocytes. Separated CD8+ cells were directly involved in tissue damage by producing necrosis of epidermis at the site of injection, with less class II antigen expression on keratinocytes, and predominantly distributed intraepidermally. CD4+ cells, located mostly in the dermal regions, induced prominent class II antigen expression on keratinocytes, but no histological changes of GVHR. High levels of TNF alpha and IFN gamma were found in the supernatant of allogeneic bMLC cultures, although when the supernatant was injected into the organotypic skin cultures, keratinocytes failed to express surface class II antigen and histologically did not show changes of skin GVHR. This study demonstrated that organotypic skin cultures can serve as a model for studying the etiology of GVHR, and indicated direct involvement of CD8+ cells in tissue damage.

Dogs undergoing rejection of unrelated, dog leukocyte antigen (DLA)-nonidentical marrow grafts show an increase in mononuclear cell counts in the peripheral blood at 1 week after transplant. Cells are of host origin and express phenotypic and morphologic characteristics of large granular lymphocytes (LGLs). LGLs from rejecting dogs suppress in vitro growth of donor marrow colony-forming units-granulocyte/macrophage (CFU-GM) and have natural killer (NK) cell activity. The current study tested whether the marrow-suppressive activity of LGLs obtained at the time of marrow graft rejection was major histocompatibility complex (MHC)-restricted. Five dogs were in the process of rejecting their DLA-nonidentical unrelated marrow grafts after conditioning with 9.2 Gy total-body irradiation (TBI). At the time of rejection, peripheral blood mononuclear cells (PBMC) were harvested. PBMC were co-cultured with marrow obtained from the original marrow transplant donor and from other unrelated dogs that were either DLA-identical or -nonidentical with the marrow donor. A statistically significant reduction of marrow donor CFU-GM was seen when compared to results with autologous effector PBMC from the marrow donor. The number of colonies with recipient effector PBMC ranged from 8 to 75% (median 29%). No suppression was seen with PBMC effectors from unrelated DLA-identical or DLA-nonidentical dogs. Similarly, significant reductions in the number of CFU-GM compared to autologous controls were seen with effector PBMC from marrow recipients and marrow target cells, both from unrelated dogs that were phenotypically DLA-identical or -nonidentical with the marrow donor. The number of colonies ranged from 6 to 68% (median 29%) and 1 to 102% (median 20%), respectively. NK activity was present at low levels in all recipients, while specific alloantigen-primed cytotoxic T cell killing by cells obtained from the five recipients yielded cytolysis of donor PBMC in only one case, suggesting that the marrow-suppressive activity was NK cell-mediated. In conclusion, PBMC from canine marrow transplant recipients undergoing rejection of DLA-nonidentical marrow grafts suppress in vitro CFU-GM growth of marrow donor cells, and this suppression is not MHC-restricted.

In this study, we examined the in vitro effect of an anti-CD44 mAb, S5, on NK function using canine PBMC as effectors. S5 enhanced NK activity in a dose-dependent and rapid fashion as did IM7, another anti-CD44 mAb that recognizes a common epitope(s) on CD44. Other anti-CD44 mAb (Hermes-1 and S3) that recognize epitopes distinct from S5 and IM7 had a variable effect on NK activity. The activation of increased killing by S5 only occurred when NK-sensitive targets were used, suggesting that lymphokine-activated killer cells were not being induced. Antibody-dependent cell cytotoxicity was not the mechanism involved in the augmentation of NK activity, nor was it Fc receptor-dependent, inasmuch as S5 F(ab')2 was able to increase NK activity. F(ab') fragments of S5 were equivalent to intact S5 in their ability to stimulate NK activity, demonstrating that cross-linking of CD44 was not a necessary component of stimulation, and that nonspecific agglutination of target and effector cells was not occurring via the two F(ab) arms. The enhancement of NK function was monocyte-independent and mediated by radioresistant cells, indicating that the antibody enhanced NK cells directly. This finding would suggest that CD44 can direct a transmembrane signal for NK cell activation.

166Holmium ethylenediaminetetramethylene phosphonic acid (166Ho-EDTMP) is a short-lived beta-emitting radionuclide complexed to an aminophosphonate ligand that we have investigated in a canine model as a potential agent for specific marrow ablation before marrow transplantation. After intravenous injections, 166Ho-EDTMP distributed principally to bone and after 24 hours the concentrations of 166Ho-EDTMP in bone were more than 200-fold higher than in any other organ. Increasing dosages of 166Ho-EDTMP led to increasingly prolonged and severe myelosuppression, but myeloablation was not achieved. Histologic examination of recovering animals suggested that the spleen may have acted as a reservoir for circulatory hematopoietic precursors. Four splenectomized animals administered 20 to 30 mCi/kg 166Ho-EDTMP without marrow transplantation died with marrow aplasia, while four splenectomized animals administered similar dosages of 166Ho-EDTMP followed by autologous transplantation recovered. The dose-limiting toxicity of 166Ho-EDTMP appeared to be marrow stromal damage resulting in myelofibrosis, which was reversible. These results suggest that 166Ho-EDTMP can be used to specifically ablate marrow function before marrow transplantation.

We established a model of canine marrow autografts after 9.2 Gy total body irradiation (TBI) to study the role of class II antigens in hematopoietic stem cell growth and differentiation. Twenty dogs were given 9.2 Gy TBI, marrow, and intravenous (IV) murine anti-class II monoclonal antibody (MoAb). Infusion of 0.6 mg/kg/d of MoAb H81.98.21, an IgG2a reactive with HLA-DR, on days 0 to 4 after TBI did not prevent initial engraftment, but dogs died with late graft failure. MoAb B1F6, an IgG2a reactive with HLA-DR + DP, had no adverse effect on engraftment, although both MoAbs detect antigens on stem cells. The critical time for the effect of MoAbs is the first 4 days after transplantation. Our findings argue against several pathogenetic mechanisms, including removal of MoAb-coated stem cells by the reticuloendothelial system (RES), canine complement-mediated cytotoxic effects on stem cells, antibody-dependent cellular cytotoxicity, and inactivation of MoAb-coated cells by dog anti-mouse antibody. To distinguish between MoAb-induced damage to microenvironment (ME)/accessory cells (AC) and late graft failure from a lack of pluripotent stem cells, three dogs were given TBI, a marrow autograft, and MoAb H81.98.21 on days 0 to 4; one, given thoracic duct cells on day 6, developed graft failure; the other two, given marrow depleted of AC by L-leucyl L-leucine o-methyl ester (Leu-Leu-OMe), had sustained grafts. Findings support the notion that originally transplanted pluripotent stem cells are no longer present on day 6 and that the ME is functional and able to support newly injected stem cells.

In marrow transplantation, radioisotope-labeled monoclonal antibodies may provide a way to selectively deliver high doses of radiation to target tissues (marrow in the case of myeloid malignancies) without significant toxicity to other normal organs. This paper describes the production and characterization of a novel monoclonal antibody, DM5, that we have developed for use in an animal model of radiotherapy targeted to the marrow. DM5 recognizes three glycosylation variants, gp19,21,23DM5, of a polypeptide core that is expressed on canine cells of the myeloid lineage, but not on lymphoid cells. The antigen recognized by DM5 is not present on most progenitor cells as determined by CFU-GM assays of DM5 positive and negative cell populations.

Pretreatment of recipients with the monoclonal antibody (MoAb) S5 facilitates engraftment of bone marrow from mismatched, unrelated donors in the canine transplantation model. In the direct comparisons reported here, the S5 glycoprotein (gp) was found to have structural homology to CD44 that in humans has been implicated in adhesive interactions of one type of effector cell, the lymphocyte. The S5 antigen and gp90Hermes-1 exhibited codistribution on canine peripheral blood cells. Both S5 and Hermes-1 (anti-CD44) MoAbs recognized 90-Kd species in radioimmune precipitations of 125I surface-labeled canine peripheral blood lymphocytes and bone marrow cells. Competitive antibody binding experiments showed that the epitope detected by S5 was distinct from that bound by Hermes-1 but overlapped with those defined by two other known anti-CD44 reagents, IM7 and Hutch-1. Sequential immunoprecipitation with S5 and Hermes-1 indicated that the two antibodies recognize the same or overlapping subsets of membrane gps. Tryptic digestion of S5 and anti-CD44 immunoprecipitates generated two major iodinated peptides of 27 and 35 Kd in both cases, a further indication of structural homology. Similarly, after N-glycanase digestion, S5 and CD44 immunoprecipitates were resolved to a single 68-Kd species. These findings suggest that CD44-mediated adhesive events may affect the fate of transplanted hematopoietic cells. The previous implications of this gp in T-lymphocyte activation and lymphocyte adhesion to endothelium thus provide useful paradigms to analyze its function in the bone marrow transplant setting.

The effect of recombinant human granulocyte-macrophage colony-stimulating factor (rhGM-CSF) on canine hematopoiesis was evaluated. rhGM-CSF stimulated granulocyte-macrophage colony formation of canine marrow depleted of accessory cells up to tenfold. Stimulation of colony formation was abrogated by anti-rhGM-CSF antiserum or heat inactivation. rhGM-CSF also stimulated in vivo canine hematopoiesis both when given as continuous i.v. infusion and as intermittent s.c. injections. Neutrophil, monocyte, and lymphocyte counts were increased three- to eightfold above controls, whereas values for eosinophils, reticulocytes, and hematocrits were not changed. Bone marrow histology after 2 weeks of treatment with rhGM-CSF showed hypercellularity with myeloid hyperplasia and left-shifted granulocytopoiesis. After discontinuation of rhGM-CSF, peripheral leukocyte counts returned to control level within 3-7 days. Platelet counts decreased rapidly after starting rhGM-CSF, to 5000-15,000 platelets/mm3, and increased within 24 h after stopping rhGM-CSF treatment, whereas marrow histology after 2 weeks of rhGM-CSF application showed the normal number and morphology of megakaryocytes.

In view of reported attempts at marrow grafting after nuclear accidents with a broad range of radiation exposures, the present study explored the total-body irradiation (TBI) conditions needed for engraftment in a canine model by using marrow from DLA-identical littermates. Previous studies have shown that such grafts are consistently successful when recipients are exposed to 920 cGy of TBI delivered at a rate of 7 cGy/min from opposing dual cobalt sources. The present TBI doses were all in the lethal range. Five dogs were administered 450 cGy; seven dogs, 600 cGy; five dogs, 700 cGy; and five dogs, 800 cGy of TBI administered at 7 cGy/min. They received a median of 3.3 x 10(8) marrow cells/kg intravenously after completion of radiation. Results showed transient allogeneic marrow engraftment in all dogs administered the lowest dose of TBI studied (450 cGy). Importantly, transient grafts permitted four of five dogs to live long enough for autologous marrow recovery to occur. At increasing radiation doses, 600, 700, and 800 cGy, the risk of graft failure lessened, with 3 of 7, 2 of 5, and 1 of 5 dogs, respectively, showing graft rejection. Fewer dogs survived with autologous marrow recovery, and more showed sustained allogeneic engraftment (4 of 7, 3 of 5, and 4 of 5 dogs, respectively). We conclude that DLA-identical littermate marrow grafts are beneficial in the setting of otherwise lethal radiation exposures, with most dogs either experiencing sustained allogeneic engraftment or surviving with autologous marrow recovery due to the extended support provided by a transient allogeneic graft.