Human gammadelta(+) T lymphocytes have in vitro graft vs leukemia activity in the absence of an allogeneic response.

Refractory acute lymphoblastic leukemia (ALL) is often incurable, and relapse rates following allogeneic bone marrow transplantation (BMT) remain high. We have reported that patients who develop increased numbers of gammadelta(+) T cells soon after BMT are significantly less likely to relapse. We now show in seven donor/recipient pairs that donor-derived Vdelta1(+)CD4(-)CD8(-)gammadelta(+) T cells are activated and proliferate in response to recipient primary ALL blasts. In addition, these cells have been shown to bind and lyse the recipient ALL blasts. Separately, gammadelta(+) T cells proliferate poorly or not at all in mixed lymphocyte culture against HLA-mismatched unrelated stimulator cells. These observations suggest that allogeneic gammadelta(+) T cells could be an effective immunotherapeutic strategy against refractory disease without the risk of graft-versus-host disease.

Partially mismatched related-donor bone marrow transplantation for pediatric patients with acute leukemia: younger donors and absence of peripheral blasts improve outcome.

PURPOSE: To extend access to bone marrow transplantation (BMT), we used partially mismatched related donors (PMRD) for pediatric patients with acute leukemia. In this report we sought to determine pretransplantation factors that might predict outcome. PATIENTS AND METHODS: Of 67 such patients, 43 had acute lymphocytic leukemia and 24 had acute myelogenous leukemia. At the time of transplantation, 41 patients were in relapse. Donors included 40 parents, 24 siblings, and three cousins. HLA disparity of two to three major antigens was detected in two thirds of the donor-recipient pairs. Conditioning therapy, including total-body irradiation and chemotherapy followed by graft-versus-host disease (GvHD) prophylaxis with partial T-cell depletion of the graft using T10B9 or OKT3, was combined with posttransplantation immunosuppression. RESULTS: Estimated probability (EP) of engraftment was 0.96 and was not affected by donor-antigen mismatch (AgMM; P =.732). EP of grades 2 to 4 acute GvHD was 0.24 and was not affected by recipient AgMM (P =.796). EP of disease-free survival was 0.26 at 3 years but improved to 0.45 when donors were younger than 30 years (P<.001). EP of relapse at 3 years was 0.41 and reduced with younger donors' age. For patients who were in relapse at the time of transplantation, absence of blasts was associated with a lower relapse rate (0.46 v. 0.84; P =. 083), similar to that of patients in remission. CONCLUSION: PMRD-BMT in pediatric leukemia resulted in high engraftment and low GvHD rates. To improve outcomes, younger donors should be sought, and clinicians should attempt to reduce peripheral blasts in patients who are in relapse.

Influence of T cell depletion method on circulating gammadelta T cell reconstitution and potential role in the graft-versus-leukemia effect.

BACKGROUND: Our laboratory previously reported that leukemia patients who developed > or = 10% gammadelta+ T cells during the first six months after receiving an anti-TCRalphabeta T-cell-depleted (TCD) graft from a partially mismatched related donor (PMRD) had a disease-free survival (DFS) advantage. These gammadelta+ T cells were V81+CD3+CD4-CD8-CD69+HLADR+ and are cytotoxic to K562 cells. METHODS: In order to determine whether the anti-alphabeta TCD regimen was associated with these findings, we compared the reconstitution of gammadelta+ T cells from patients who received TCD PMRD grafts using the anti-TCRc4 MAb TIOB9-1A31 (previously reported) with similar patients who received grafts using the anti-CD3 MAb OKT3. RESULTS: Increased cytotoxic Vdelta1+ T cells were seen in 10 of 43 T10B9 TCD patients compared to 7 of 100 in the OKT3 TCD group (23% versus 7%, p = 0.010). T10B9 patients with increased gammadelta+ T cells also exhibited a higher range of increased gammadelta+ T cells and the length of time the gammadelta+ T cells remained high was longer when compared to OKT3 patients. Patients with increased gammadelta+ T cells whose grafts were T-cell depleted with T10B9 showed a significant decrease in relapse (p = 0.038). Similar rates and reduction in relapse were seen in OKT3 TCD patients, although significance was not reached due to the small number of patients with increased gammadelta+ T cells. Estimated 3 year disease-free survival was significantly improved in T10B9 patients with increased gammadelta+ T cells (0.79 versus 0.31, p = 0.009), a trend also seen in OKT3 patients (p = 0.091). DISCUSSION: These observations indicate that Vdelta1+CD4-CD8-cytotoxic T cells are associated with lower relapse rates and improved survival, and thus may have a role in a graft-versus-leukemia effect.

Donor leukocyte infusion for treatment of graft rejection post partially mismatched related donor bone marrow transplant.

Graft rejection following bone marrow transplantation is more common in patients who receive their grafts from alternative donors and whose marrow is T cell depleted. Rejection in these patients is mediated by persistent host cells that interfere with successful establishment of donor-derived hematopoietic recovery. We describe a patient with chronic myelogenous leukemia in accelerated phase who rejected a T cell-depleted bone marrow graft, 2 months following partially mismatched related donor bone marrow transplant. Unmanipulated peripheral blood donor leukocyte infusion, without additional chemotherapy or immunosuppressive therapy resulted in complete hematopoietic recovery. Cytogenetics and RFLP demonstrated hematopoietic donor chimerism. The patient did not develop graft-versus-host disease.

Phenotypic and functional reconstitution of peripheral blood lymphocytes following T cell-depleted bone marrow transplantation from partially mismatched related donors.

Myeloablative chemotherapy followed by transplantation of a T cell-depleted bone marrow graft from a partially mismatched related donor provides a potentially curative option for patients with leukemia and other disorders of hematopoiesis, although the patient is faced with a period of sustained immunodeficiency as well as pharmacologic immunosuppression as a result of prophylaxis against graft-versus-host disease. Thirty patients who received one to three antigen T cell-depleted mismatched grafts were evaluated for immune reconstitution. The percentage and numbers of cells expressing lymphocyte subset antigens were determined by flow cytometry at 14, 28, 60, 100, 180, 270 and 365 days post-BMT and at 6 month intervals thereafter. Lymphocyte reconstitution was characterized by the early appearance of natural killer cells and a low percentage of both T and B cells. During the first year after BMT, the number of NK cells remained constant while T and B cells gradually returned to normal numbers and proportions. Response to the lymphocyte mitogen phytohemagglutinin returned to normal over the course of 2 years, while the response to concanavalin A was slightly depressed and the response to pokeweed mitogen became supranormal at about 1.5 years and continued to increase. These data suggest the need for long-term immunophenotypic monitoring as well as prolonged infection surveillance and prophylaxis.

Use of partially mismatched related donors extends access to allogeneic marrow transplant.

Most patients requiring allogeneic bone marrow transplant (allo-BMT) do not have an HLA-matched sibling donor. A phenotypically matched unrelated donor graft has been made available for approximately 50% of Caucasians and less than 10% of ethnic and racial minorities in need. However, almost all patients have a readily available partially mismatched related donor (PMRD). We summarize our experience with 72 patients who ranged from 1 to 50 years of age (median, 16 years) and who were recipients of a PMRD allo-BMT from haploidentical family members following conditioning therapy using total body irradiation (TBI) and multiagent, high-dose chemotherapy. T-cell depletion and post-BMT immunosuppression were combined for graft-versus-host disease (GVHD) prophylaxis. The probability of engraftment was 0.88 at 32 days. Six of 10 patients who failed to engraft achieved engraftment after secondary transplant. Grade II to IV acute GVHD was seen in 9 of 58 (16%) evaluable patients; extensive chronic GVHD was seen in 4 of 48 (8%) evaluable patients. There was a statistically significant difference in 2-year survival probability between low-risk and high-risk patients (0.55 v 0.27, P = .048). Prognostic factors that affected outcomes in multivariate analysis were (1) a lower TBI dose and 3-antigen rejection mismatch decreased stable engraftment (P = .005 and P = .002, respectively); (2) a higher T-cell dose increased acute GVHD (P = .058); (3) a higher TBI dose increased chronic GVHD (P = .016); and (4) a high-risk disease category increased treatment failure from relapse or death (P = .037). A PMRD transplant can be performed with acceptable rates of graft failure and GVHD. Using sequential immunomodulation, the disease status at the time of transplant is the only prognostic factor significantly associated with long-term successful outcome after PMRD allo-BMT. When allogeneic rather than autologous BMT is indicated, progression in disease status before transplant can be avoided using a PMRD with equal inclusion of all ethnic or racial groups.

Characterization of acute bone marrow graft rejection in T cell-depleted, partially mismatched related donor bone marrow transplantation.

The purpose of this study was to characterize the phenotype and clonality of the T cell population in patients who experience acute rejection (AR) following bone marrow transplantation (BMT) from a partially mismatched related donor (PMRD). Phenotypic analysis was performed using flow cytometry, assignment of donor/host lineage by cytogenetics or HLA-specific flow cytometry, and analysis of the T cell receptor (TCR) by reverse-transcriptase polymerase chain reaction (RT-PCR). We have previously reported the initial appearance in the blood of AR patients of host CD8+brightCD3low T cells that progressively express increasing amounts of CD3+ cells. We now report that this cell population can differentiate into either a cytotoxic T cell phenotype (CD3+CD8+HLA-DR+CD57-) usually associated with AR of grafts from matched unrelated donors or a suppressor T cell phenotype (CD3+CD8+CD57+HLA-DR-) usually associated with AR of grafts from matched sibling donors. Analysis of the TCR V beta subsets from two patients revealed sorted host CD3+CD8+ cells (purity 90-95%) from the first patient to express V beta 18 almost exclusively. In a second patient with late rejection (55 days post-BMT), the CD3+CD8+ cells were predominantly restricted to V beta 1, 5.1, 7, 9, and 18. Although CD3+CD8+ T cells are known to be associated with AR, cytotoxic and suppressor lineages in AR from the same type of BMT and clonal distribution of T cells in AR have not been reported. Preliminary results suggest that V beta expression in AR of PMRD grafts is restricted and host T cell phenotype may vary. Further studies will investigate whether specific mismatches correlate with specific V beta usage and/or host T cell phenotype.

Evidence that the antigen receptors of cytotoxic T lymphocytes interact with a common recognition pattern on the H-2Kb molecule.

Recognition of class I MHC antigens involves interaction between TCRs of cytotoxic T lymphocytes (CTL) and the two alpha helices of MHC molecules. Using a combined panel of H-2Kb mutants selected by either a CTL clone or MAbs, we have shown evidence that the TCRs of 59 Kb-specific CTL clones shared a common binding pattern on the H-2Kb molecule. Mutations of amino acid residues at the C-terminal regions, but not the N-terminal regions, of the alpha helices abrogated the recognition by the majority of the clones. The data suggests that TCRs predominantly recognize the class I MHC molecule with an orientation that is parallel to the beta-pleated strands and diagonal to the alpha helices.

Characterization of a novel human corneal endothelial antigen.

The antigenic composition of the human corneal endothelium, a cellular layer essential for maintaining corneal function, has not been well characterized. A novel corneal endothelial antigen was identified by generating a monoclonal antibody (MAb) against normal human corneal endothelial cells. This MAb, designated 2B4.14.1, reacted strongly by immunoperoxidase staining with the endothelium of corneas from all human donors tested but not with other corneal components, including epithelium and stroma. Positive immunohistologic reactions of 2B4.14.1 with several other human tissues, including kidney (parietal epithelium of Bowman's capsule, proximal convoluted tubule, ascending limb of Henle's loop, and distal convoluted tubule), glandular epithelia of numerous organs, and mesothelial linings of several thoracic and abdominal viscera, also were observed. One of the renal antigens recognized by 2B4.14.1 was identified as Tamm-Horsfall glycoprotein (THGP), based on the ability of the antibody to recognize THGP in western immunoblots and the abrogation of immunohistologic reactivity of the antibody by preincubation with purified THGP. These findings raise the possibility that the human cornea expresses a molecule with homeostatic properties similar to those ascribed to THGP. However, it is unlikely that the corneal antigen recognized by 2B4.14.1 is conventional THGP; a MAb specific for THGP did not react with corneal endothelium.

Cytotoxic T lymphocyte recognition of novel allodeterminants expressed on in vitro selected H-2Kb mutants.

In the present study, in vitro derived H-2Kb mutants have been examined by alloreactive CTL. Two mutants, R8.24 and R8.246, have been shown to express novel determinants detected by CTL generated against some but not all in vivo derived Kb mutants. BDF1 anti-bm3, anti-bm11, anti-bm19, anti-bm23, and anti-bm6 CTL populations lyse the two R8 variants. The novel determinants expressed on the R8 mutants detected by the bm3 and bm23-specific CTL appear to differ from the determinant recognized by the bm6-specific CTL. No new serologically defined determinants were detected on any of 18 independent R8 variants. However, these results do not rule out the existence of new determinants which could be recognized by antibodies. Finally, the relationship between the T cell recognition of the in vivo and in vitro derived mutants and their use in understanding the structure/function relationships between the immune response and class I Ag based on recent crystallographic analyses is discussed.

Evidence that multiple residues on both the alpha-helices of the class I MHC molecule are simultaneously recognized by the T cell receptor.

Single amino acid substitutions at nine different positions on the H-2Kb molecules from in vitro-mutagenized, immunologically altered, somatic cell variants were correlated with their patterns of recognition by monoclonal antibodies (MAbs) and allogeneic cytotoxic T lymphocyte (CTL) clones. While MAbs were found to detect spatially discrete, domain-specific sites, CTLs interacted simultaneously with multiple residues on the alpha 1 and alpha 2 domains of the Kb molecule. The computer graphic three-dimensional Kb model structure showed that, of the seven CTL-specific residues analyzed, six residues were located on the alpha-helical regions of the two domains. Every CTL clone was found to interact with a distinct pattern of residues composed of a specific subset of the CTL-specific residues.

Molecular loss variants of the murine major histocompatibility complex: nonexpression of H-2K antigens associated with marked reduction in H-2K mRNA as determined by oligonucleotide hybridization analysis.

Somatic cell variants of the murine major histocompatibility complex were isolated to study the molecular features required for H-2K gene expression. In vitro selection was performed on a heterozygous [H-2b (Kb,Db) X H-2d (Kd, Dd, Ld)] pre-B lymphoblastoid cell line (R8) for variants that had lost membrane expression of the H-2Kb gene product. Analysis of a number of independently isolated variant cell lines by cytofluorometry with monoclonal antibodies to the Kb, Kd, Db, and Ld antigens revealed a variety of H-2 phenotypes. Variants were classified as either molecular loss for those that had lost K antigen expression only or as haplotype loss for those that no longer expressed the entire H-2b haplotype (i.e., negative for Kb and Db). DNA hybridization analysis with a K gene-specific oligonucleotide indicated that the Kb gene was present in all of the molecular loss variants, suggesting that gene deletion was not responsible for the loss of Kb antigen expression. In contrast, the Kb gene was not detected in haplotype loss variants. For analyzing the mutants at the RNA level, hybridization with H-2-specific synthetic oligonucleotides provided a definitive procedure to identify specific class I gene transcripts in the H-2 heterozygous cell lines. Such analyses were performed on the molecular loss variants and revealed that in a subset of variants (R8.2, R8.96, R8.116, R8.178) there was an absence of Kb mRNA. Kd mRNA was identified for all but one (R8.2) of the Kb mRNA-deficient cell lines, a finding consistent with the serotype assigned to each. Transcription of the linked Db gene was normal. These data demonstrate that a specific alteration for the K gene in several independently selected cell lines gave rise to the altered H-2 phenotype. Such variants offer the potential to analyze the properties of the mammalian genome responsible for controlling expression of individual members of the major histocompatibility complex multigene family.

An approach to the study of structure-function relationships of MHC class I molecules: isolation and serologic characterization of H-2Kb somatic cell variants.

Somatic cell variants expressing an altered antigenic form of the H-2Kb molecule were isolated for the purpose of performing structure-function analysis of a class I MHC molecule. Over 25 independently isolated variants were derived from an Abelson virus transformed pre- B cell line (R8) by mutagenesis with ethyl methane sulfonate or ethyl nitrosourea. Negative selection was performed by complement-dependent cytotoxicity with anti-H-2Kb monoclonal antibodies subsequently followed by positive selection to separate the H-2Kb surface negative variants from structural variants. Biochemical characterization of a random selection of three independent variants indicated that the variant H-2Kb molecule was present in normal amounts in lysates, and was unchanged in size. Cytofluorometric analysis with the use of a panel of seven monoclonal antibodies against H-2Kb indicated that all of the variants had lost one or more alloantigenic determinants (monoclonal antibody binding sites). For these variants, the pattern of monoclonal antibody loss of recognition suggested that antibody defined alloantigenic determinants appear to be discretely localized to a single domain, either the alpha 1 or the alpha 2 domain, of the H-2Kb molecule. In contrast, CTL recognition of the Kb molecule of these variants depends on involvement of both alpha 1 and alpha 2 domains as shown in the companion paper.

Somatic cell variants express altered H-2Kb allodeterminants recognized by cytolytic T cell clones.

The present studies have made use of in vitro derived H-2Kb mutants to analyze the fine specificity of alloreactive cytotoxic T lymphocytes (CTL). The variants were derived by negatively selecting mutagenized tumor cells with a monoclonal anti-H-2Kb antibody and positively selecting for residual cells expressing serologically altered H-2Kb molecules. Details of this procedure are described in the companion paper. Selected populations of bulk alloreactive and cloned CTL were examined for recognition of the variants. In contrast to the serologic findings presented in the companion paper, there does not appear to be a correlation between the monoclonal antibody used to select the R8 variant and the CTL specificities recognized. In several instances, CTL clones could discriminate between variants having identical serologic profiles. Therefore, it would appear that the CTL have a large repertoire of allorecognition, even when generated across a mutant anti-Kb combination reflecting only a few amino acid differences. In addition, a diverse set of epitopes can be recognized on the Kb molecule. Finally, in some instances a change in what would appear to be a single amino acid resulted in a profound alteration of CTL recognition even though the Kb mutant molecule expressed limited serologic changes. These results support the idea that small changes in the H-2Kb molecule can have dramatic effects on CTL even though there are relatively little effects on serologic recognition of the target molecule.

Somatic cell variants of H-2Kb: a point mutation in the first extracellular domain results in altered immune recognition.

A cell-surface-associated variant H-2K product was expressed by an Abelson virus-induced pre-B-cell line after chemical mutagenesis with ethyl methane sulfonate. The variant cell line (R8.313) was previously demonstrated to have altered allodeterminants in Kb as demonstrated by both Kb-specific monoclonal antibody binding and alloreactive cytotoxic T lymphocyte (CTL) cytolysis. The mutant H-2Kb gene from R8.313 was cloned and characterized in detail. DNA sequence analysis of the region of the gene corresponding to the three extracellular domains identified a single point mutation resulting in a leucine-to-phenylalanine substitution at amino acid residue 82. The site of mutation within the alpha 1 domain was confirmed by oligonucleotide hybridization analysis. Mouse L-cell fibroblasts transfected with the mutant gene were recognized with the same monoclonal antibody binding and CTL lytic pattern as the R8.313 cell line, confirming that the altered phenotype of the mutant cell line was due to a point mutation in the H-2Kb gene. These data further extend the hypothesis that the region of amino acid residues 70-90 in the alpha 1 domain is important in the formation of both antibody and CTL-defined recognition structures on major histocompatibility complex class I molecules.

Differentiation of astrocytes and oligodendrocytes from germinal matrix cells in primary culture.

Primary cultures derived from neonatal rat forebrain grow almost entirely as glial cultures, with a large astrocytic preponderance and smaller numbers of oligodendrocytic cells. Although both astrocytic and oligodendrocytic characteristics are acquired in vitro, the origins of both types of glia in primary cultures have not been determined. We tested the hypothesis that glia differentiate in vitro from immature neuroectodermal cells by following the fate of germinal zone cells in primary cultures. A monoclonal antibody that binds GD3 ganglioside was used as a marker for cells of the subventricular zone (SVZ), since antibody binding in newborn rat forebrain could be detected by immunofluorescence only in the SVZ of newborn rats (Goldman et al., 1984). We followed the expression of glial fibrillary acidic protein (GFAP), an astrocytic marker; galactocerebroside (GC), an oligodendrocytic marker; and GD3 during the first several weeks of culture. Both GFAP and GC expression were first detected in cells that bound the GD3 antibody. Astrocytes developed during the first week in vitro; eventually, they lost the ability to bind the GD3 antibody and most became GD3-/GFAP+ cells. In high-density cultures, a population of small cells that resided on top of the astrocytic monolayer retained GD3 expression. GC-antibody binding was first observed in these cells of the upper layer, although it was not readily apparent until the second week of culture. Few GC+ cells were seen in cultures grown at low density, however.(ABSTRACT TRUNCATED AT 250 WORDS)

Secondary structure of the murine histocompatibility alloantigen H-2Kb: relationship between heavy chain, beta 2-microglobulin, and antigenic reactivity.

The far-ultraviolet circular dichroism (CD) spectra of the extracellular portion (papain-cleaved fragment) of the histocompatibility antigen H-2Kb and its noncovalently associated components, heavy chain and beta 2-microglobulin (beta 2m), indicate that the antigen is highly structured, containing about 30% alpha-helix, 41% beta-sheet, and 29% random coil. Separation of beta 2m from the heavy chain produced a decrease in heavy chain alpha-helix and beta-sheet structure which correlated with a loss of alloantigenic reactivity. Reconstitution of the heavy chain-beta 2m complex resulted in an increase in secondary structure which was greater than the sum of the free chains and the recovery of considerable alloantigenic reactivity. This suggests that some of the secondary structure and much of the alloantigenic reactivity may depend on conformation associated with the binding of beta 2m to heavy chain. A prediction of heavy chain secondary structure based on Chou-Fasman analysis of the primary amino acid sequence agreed with results from CD measurements and suggested that the segments of alpha-helix and beta-sheet structure are distributed throughout the molecule.

Comparison of the cloned H-2Kbm1 variant gene with the H-2Kb gene shows a cluster of seven nucleotide differences.

The gene for the H-2K class I antigen of the bm1 variant was cloned and analyzed at the DNA level and compared with the previously cloned parent B6/Kh gene. Sequence determination and comparative restriction endonuclease studies indicate that Kbm1 is derived from the Kb gene. Seven nucleotide changes within a 13-nucleotide stretch distinguish the mutant from the parent gene and result in amino acid differences at positions 152, 155, and 156 in the antigen. The data confirm previously reported changes at amino acid positions 155 and 156 (arginine to tyrosine and leucine to tyrosine, respectively) and extend the altered region to include two nucleotides encoding a glutamate to alanine substitution at amino acid 152, a change not detected by the protein studies because of limitations of the methods used. The DNA sequence encoding this region of the Kbm1 glycoprotein is identical to the DNA sequence of at least one other known class I gene in the mouse, a finding consistent with the hypothesis that the mutation was not a random event but may be the result of a block transfer of information by a copy mechanism analogous to gene conversion. As the sequence analysis of the coding region for the first 273 amino acid residues shows identity between parent and mutant except for the seven nucleotide changes, all variant-parent functional differences must depend only on the cluster of three amino acid differences in the second domain of the Kb glycoprotein.

In vivo antigenic modification of tumor cells. I. Introduction of murine leukemia virus antigens on non-virus-producing murine sarcomas.

Murine oncovirus antigens represent excellent targets for immune recognition, and virus-associated tumors are generally susceptible to various immunotherapy protocols. Virus-negative tumors, however, are nonimmunogenic and refractory to immunologic control. Therefore, the feasibility of the introduction of antigens onto non-virus-expressing tumors in situ in inbred C57BL/6J mice by systemic administration of nononcogenic murine retroviruses was investigated. Two classes of murine fibrosarcomas were studied: a 3-methylcholanthrene-induced fibrosarcoma syngeneic to C57BL/6 mice (MCA-FS) and a Harvey murine sarcoma virus-transformed, nonproducer fibrosarcoma syngeneic to C57BL/6 mice (H-NP). Both were found to be devoid of infectious ecotropic murine leukemia virus (MuLV) or MuLV antigens. A single dose of Friend murine leukemia virus (F-MuLV) was used to superinfect MCA-FS- and H-NP-induced tumors in vivo and converted these tumors to a highly productive, virus-positive state. In vivo superinfected tumors were indistinguishable from their preinfected counterparts by competition radioimmunoassays for the virion's major envelope glycoprotein, gp71, and its group-specific antigen, p30, and by assays for infectious virus. Analysis of virus from tumor extracts proved that the antigenic specificity of the superinfected tumor was provided by F-MuLV administered systemically to the animals. Finally, an immunoperoxidase technique, applied to tumor cross sections, demonstrated the uniform appearance of viral antigens in the superinfected tumors.