Search

Allogeneic hematopoietic stem cell transplantation (allo-SCT) followed by donor lymphocyte infusion (DLI) is a potential curative treatment for patients suffering from a hematological malignancy. Efficacy is attributed to the graft-versus-tumor (GVT) response, during which engrafted donor T cells become activated by recipient minor histocompatibility antigens (MiHA) presented on dendritic cells (DC). Subsequently, these activated T cells expand, acquire effector functions and kill MiHA-positive tumor cells. However, persistence and recurrence of malignant disease is often observed, indicating that insufficient GVT immunity is induced. This imperfect alloreactive response is probably due to insufficient numbers of MiHA-specific effector T cells and/or defective antigen-presentation and costimulation. Therefore, adoptive transfer of potent ex vivo-generated MiHA-specific T cells, restricted to the hematopoietic system, would boost the GVT-effect without increasing the risk for GVHD. Although successful in vitro induction of MiHA-specific CD8+ T cells from naive precursors has been reported, the resulting antigen-experienced T cell population consist of fully differentiated effector-memory T cells (TEM). Over the past years it has been described that this T cell subset is not the most potent memory subset in anti-tumor responses in vivo following T cell transfer. In this regard, the less-differentiated memory subset called stem cell memory T cells (TSCM) with superior in vivo expansion, self-renewal capacity and plasticity to differentiate in potent effectors would generate a stronger GVT response. In this study, we aimed to investigate the in vivo availability and ex vivo generation of TSCM-like MiHA-specific T cells as additive treatment option for allo-SCT patients. First, we investigated whether in allo-SCT patients MiHA-specific T cells could be detected with a TSCM phenotype defined by the expression of CD45RO, CCR7, CD27 and CD95. Though TSCM cells could be clearly detected within CMV-specific CD8+ T cells in allo-SCT patients, similar to healthy controls, no MiHA-specific TSCM cells could be detected. This emphasises the need for more potent adoptive MiHA-specific T cell therapy following allo-SCT. Therefore, we next explored the possibility of generating TSCM-like CD8+ T cells by interfering with the Akt signalling pathway. Emerging findings indicate that the differentiation program of CD8+ T cells is dictated by the strength and duration of AKT activity. Therefore, we explored whether the pharmacological inhibition of this signaling pathway could results in the generation of TSCM-like CD8+ T cells. We stimulated CCR7+CD45RA+ naive CD8+ T cells with CD3/CD28 beads plus IL-2, IL-7 and/or IL-15 in the presence an Akt inhibitor. Interestingly, CD8+ T cells in these Akt-cultures were inhibited in their differentiation stage, expressing higher levels of CD45RA and CCR7 compared to controls. In addition, expression of CD95, IL2Rβ, and IL7Rα was also elevated confirming the TSCM-like phenotype. Although proliferation of the Akt-inhibited CD8+ T cells was decreased as shown by less PBSE dilution, expansion could be significantly preserved. Next, we investigated whether the established culture conditions could be used to generate MiHA-specific TSCM-like cells. Therefore, CD8+ T cells from MiHA-negative donors were primed using autologous MiHA peptide-loaded moDCs in the presence of the Akt-inhibitor. Interestingly, MiHA-specific T cell priming could be induced, consisting of mainly TCM and TSCM-like cells compared to almost entirely TEM cells in the control setting. Akt-inhibited MiHA-specific T cells showed higher expression of CCR7, CD45RA, CD62L, CD28, CD95, and IL7Rα. Importantly, for the Akt-inhibited MiHA-specific T cells, proliferation was reserved, resulting in robust proliferation capacity during restimulation after removal of the Akt-inhibitor. The resulting TEFF cells were highly functional, showing capacity to degranulate and produce IFNγ upon peptide restimulation. In conclusion, by inhibiting the Akt-pathway, in vitro CD8+ T cell differentiation can be reduced. Therefore, Akt signalling inhibition can be exploited for generating TSCM-like MiHA-specific T cells in adoptive immunotherapy after allo-SCT. Disclosures: No relevant conflicts of interest to declare.

Natural killer (NK) cell-based immunotherapy is a promising adjuvant, relatively non-toxic therapy approach for AML. However, further improvement of NK cell-based therapy is needed to increase the clinical effect. In this regard, NK cells generated ex vivo from hematopoietic progenitor cells (HPC) may have significant clinical benefits over enriched NK cells from adult donors, including the ability to choose an appropriate killer-cell immunoglobuline-like receptor (KIR)-ligand or KIR B haplotype alloreactive donor, as well as the capacity to reach high therapeutic dosages. Previously, we reported a GMP-compliant, cytokine/heparin-based culture protocol for the ex vivo generation of highly active NK cells from CD34+ HPC isolated from cryopreserved umbilical cord blood (UCB) units. Expansion in closed, large-scale bioreactors yields a clinically relevant dose of NK cells with high purity and cytolytic activity against AML cells in vitro. Currently, a clinical phase I trial with these HPC-NK cells is ongoing in our hospital. Trafficking studies in NOD/SCID/IL2Rgnull (NSG) mice demonstrated that these HPC-NK cells migrate to the bone marrow (BM) as well as to lymphoid organs where in vivo expansion and maturation can take place. Analysis of the chemokine receptor expression profile of UCB-NK cells matched in vivo findings. Particularly, a firm proportion of UCB-NK cells functionally expressed CXCR4, what could trigger BM homing in response to its ligand CXCL12. In addition, high expression of CXCR3 and CCR6 supported the capacity of UCB-NK cells to migrate to inflamed tissues via the CXCR3/CXCL10-11 and CCR6/CCL20 axis. Importantly, a single HPC-NK cell infusion combined with supportive IL-15 administration was shown to efficiently inhibit growth of K562 leukemia cells implanted in the femur of NSG mice, resulting in significant prolongation of mice survival. Furthermore, we investigated whether modulation by the DNA methyltransferase (DNMT) inhibitors Azacytidine (Aza) and Decitabine (Deci) could further potentiate the antileukemic effect of HPC-NK cells against AML cells. In concordance with previous reports, we observed a dose-dependent effect of Aza and Deci on the growth of the AML cell lines THP1 and KG1a. In subsequent NK cell killing assays, we used clinical relevant low drug concentrations to pre-treat AML cells that did not affect HPC-NK cell viability and cytolytic function. Interestingly, increased killing of pre-treated THP1 and KG1a cells by HPC-NK cells could be observed, which was correlated with an increase in the NKG2D ligand ULBP2, the DNAM-1 ligands CD112 and CD155 as well as TRAIL-R2. Notably, maintenance of low-dose DNMT inhibitors during the KG1a/NK co-culture resulted in pronounced AML growth inhibition. To examine the effect of DNMT inhibitors in vivo, THP1.LucGFP-bearing NSG mice were treated with increasing dose of both agents, which were administered according to current standard protocols applied in humans. Data indicated that treatment with Aza or Deci at dosage equivalent in human to 12.5 and 5 mg/m2 respectively was well tolerated with minimal and/or transient weight loss, and efficiently reduced the progression of THP-1.LucGFP cells in vivo. Currently, we explore whether HPC-NK cells and DNMT inhibitors can work together to combat AML in our xenograft models. These preclinical studies may provide a rationale to investigate the possible additive and/or synergistic anti-AML effects of adoptive HPC-NK cell transfer in combination with these DNMT inhibitors in AML patients. Disclosures: Tordoir: Glycostem Therapeutics: Employment. Spanholtz:Glycostem Therapeutics: Employment.

Abstract 3242 Although the importance of the bone marrow (BM) in hematopoiesis is well known, its function in adaptive immune responses has only recently been acknowledged. Currently it is known that the BM contains fully functional CD4+ and CD8+ T cells that can engage in both primary and secondary immune responses. Interestingly, most of these T cells belong to the memory T cell lineage, identifying the BM as one of the largest memory T cell reservoirs in the body. Since not much is known about the trafficking of BM T cells, we compared the homing phenotype and function of T cell subsets in the BM, blood, spleen and peripheral lymph nodes (pLN). In addition, we determined the expression of chemokine mRNA and protein levels in the BM and other lymphoid organs. We confirmed that at least 80% of the CD4+ and 60% of the CD8+ BM T cells have a memory phenotype, and that most CD4+ T cells belong to the effector memory lineage, while the CD8+ population predominantly consists of central memory T cells. Most BM T cells expressed the chemokine receptor CXCR3, the adhesion molecules P-selectin glycoprotein ligand 1 and VLA-4, and increased levels of CD44 and LFA-1, as compared to T cells from the spleen. In addition, L-selectin was absent from most CD4+ BM T cells, but present on virtually all CD8+ T cells. Notably, the percentage of CXCR3+ T cells within the effector memory and central memory subsets from BM was higher than within the same subsets from pLN. Furthermore, BM contained significant mRNA levels of the CXCR3 ligands CXCL9, CXCL10 and CXCL11. An in vivo migration assay using a mixture of fluorescent-labeled T cells from CXCR3-deficient mice and control mice indicated however that during homeostasis CXCR3 does not play a major role in BM entry or retention. These data suggest that CXCR3 expressed by memory T cells is rather involved in BM exit, than in BM entry. Indeed, we observed that, as compared to control mice, CXCR3−/− mice contained significantly more CD4+ and CD8+ T cells in their BM. Additional in vitro assays demonstrated that CD4+ and CD8+ BM T cells migrated vigorously in response to CXCL9 and CXCL10, generally released in high concentrations during inflammation. Finally, we demonstrate that CXCR3−/− effector/effector memory T cells, but not wild type T cells, accumulate in the BM of mice infected with lymphocytic choriomeningitis virus. Altogether, these data demonstrate that the BM is a major reservoir of memory T cells that employ CXCR3 to quickly respond to chemotactic signals from inflamed tissues. Disclosures: No relevant conflicts of interest to declare.

Allogeneic stem cell transplantation (SCT) in combination with donor lymphocyte infusion (DLI) is an experimental treatment for patients with metastatic solid tumors. The therapeutic efficacy is attributed to the graft-versus-tumor (GVT) response during which donor-derived T cells eliminate malignant cells via recognition of minor histocompatibility antigens (MiHA). To reduce accompanying GVHD, it is crucial to identify MiHA which are selectively expressed on hematopoietic cells and solid tumor cells. Previously, we identified a hematopoietic cell-restricted MiHA, designated LRH-1, which is presented by HLA-B7 and encoded by the P2X5 purinergic receptor gene (J. Clin. Invest.2005:115:3506–3516). Here, we report that LRH-1, in addition to its hematopoietic cell-restricted expression, is aberrantly expressed on epithelial tumor cell lines. We observed that P2X5 mRNA is significantly expressed in 14 out of 42 (33%) solid tumor cell lines tested by real-time quantitative RT-PCR. We detected P2X5 transcripts in 3 out of 11 renal cell carcinoma cell lines, 2 out of 4 melanoma cell lines, 3 out of 7 colorectal carcinoma cell lines, 4 out of 10 brain tumor cell lines and 2 out of 10 breast cancer cell lines. To determine whether P2X5 mRNA expression in solid tumor cell lines results in susceptibility to lysis by LRH-1-specific CTL, we performed flow cytometry-based cytotoxicity assays using P2X5-expressing tumor cell lines. Based on LRH-1 genotyping analysis, we selected six solid tumor cell lines for the cytotoxicity studies. Remarkably, LRH-1-specific CTL efficiently lysed and inhibited the growth of DAOY brain tumor cells up to 3 days of co-culture. The renal cell carcinoma cell lines SKRC-33 and SKRC-18 and the melanoma cell line BLM were also susceptible to LRH-1 CTL-mediated lysis, although less effectively. However, pre-incubation of these tumor cell lines with IFNγ and TNFα significantly increased the susceptibility to LRH-1-specific CTL and resulted in complete target cell lysis. Furthermore, these cytokine-stimulated cell lines induced higher levels of CTL degranulation as determined by CD107a staining. No cytotoxicity was observed against LRH-1-negative FM3 melanoma and SKRC-24 renal cell carcinoma cell lines. These findings illustrate that the GVT reactivity observed in solid tumors after allogeneic SCT may be selectively enhanced by LRH-1-specific immunotherapy.

Allogeneic hematopoietic stem cell transplantation (HSCT) is a potent treatment for patients with hematological malignancies. The therapeutic efficacy is attributed to the graft-versus-tumor (GVT) response during which donor-derived cytotoxic T lymphocytes (CTL) eliminate malignant cells of the recipient. Minor histocompatibility antigens (MiHA) are the major targets of the GVT response, and expansion of MiHA-specific CTL has been shown to coincide with tumor remission following SCT. Unfortunately, GVT response is often accompanied by graft-versus-host disease (GVHD) causing severe damage to skin, liver and gut. Therefore, it would be highly beneficial to direct GVT immunity to MiHA that are selectively expressed by “malignant” hematopoietic cells. Recently, we identified a novel lymphoid lineage-restricted MiHA, designated LRH-1, which is derived from a frame shift polymorphism in the P2X5 purinergic receptor protein (J. Clin. Invest.2005:115:3506–3516). Here, we examined by real-time quantitative RT-PCR mRNA expression of P2X5 in patient samples and cell lines from numerous lymphoid malignancies. We observed that P2X5 mRNA is highly expressed in tumor cells from all stages of lymphoid development. Furthermore, we demonstrated that LRH-1−specific CTL efficiently lyse LRH-1+ tumor cells and cell lines of lymphoid origin. These findings illustrate that LRH-1 is an attractive target for specific immunotherapy after allogeneic HSCT. A potentially efficient strategy is the application of MiHA-loaded dendritic cells (DC) to boost MiHA-specific T cell responses already primed in vivo after HSCT. To facilitate efficient presentation of LRH-1 by DC, we have optimized RNA electroporation of DC using in vitro-transcribed mRNA. We observed that RNA electroporation of mature DC resulted in a high yield of viable cells, high expression of co-stimulatory molecules, no loss of migratory capacity towards lymph node-specific chemokines, and high protein expression of the introduced antigens. Furthermore, we demonstrated that RNA-electroporated DC display long-lasting peptide presentation to LRH-1− specific CTL and induce proliferation of LRH-1− specific effector-memory CTL ex vivo. These data will be of importance in designing LRH-1− based immunotherapy in transplanted patients with lymphoid malignancies.

Allogeneic stem cell transplantation (SCT) is a specialized form of immunotherapy for treating patients with hematological malignancies. The curative potential is attributed to the graft-versus-tumor (GVT) response during which donor-derived cytotoxic T lymphocytes (CTL) eliminate malignant cells of the recipient. Minor histocompatibility antigens (MiHA) are the major targets of the GVT response, and expansion of MiHA-specific CTL has been shown to coincide with tumor remission following SCT. Recently, we identified a novel hematopoietic cell-restricted MiHA, designated LRH-1, which is presented by HLA-B7 and encoded by the P2X5 purinergic receptor gene (J. Clin. Invest.2005:115:3506–3516). Interestingly, tetramer analysis showed a direct association between in vivo expansion of LRH-1-specific CD8+ T cells and the disappearance of Bcr-Abl positive tumor cells in the CML patient from whom LRH-1-specific CTL was originally isolated. In addition, we detected in vivo expansion of LRH-1-specific CTL in an AML patient who was in clinical remission without GVHD. Furthermore, we demonstrated that P2X5 mRNA is significantly expressed in leukemic CD34+ progenitor cells from most CML as well as AML patients. These findings indicate a role for LRH-1 in inducing GVT immunity against myeloid leukemic progenitor cells. Here, we investigated the ex vivo responsiveness of myeloid leukemic CD34+ progenitor cells to LRH-1-specific CTL. First, we addressed this question using the CD34+ KG1 cell line which is positive for LRH-1. By using a CFSE-based survival assay we demonstrated that KG1 cells stably transfected with HLA-B7 could be efficiently lysed by LRH-1-specific CTL. Next, we determined responsiveness of purified CD34+ progenitor cells from HLA-B7+ CML patients to LRH-1 CTL-mediated killing. In the CFSE-based survival assay as well as a hematopoietic progenitor cell inhibition assay we showed that LRH-1-specific CTL efficiently recognize and kill CD34+ progenitor cells from LRH-1+ CML patients. In contrast, LRH-1-specific CTL did not inhibit the proliferation of CD34+ progenitor cells from LRH-1- CML patients. These findings illustrate that the P2X5-encoded LRH-1 antigen is an attractive target for adequate eradication of myeloid leukemic progenitor cells after allogeneic SCT.

Allogeneic stem cell transplantation is a potent treatment for patients with leukemia. This therapeutic effectiveness is largely attributed to the graft-versus-leukemia (GVL) response during which alloreactive donor CD8+ T cells eradicate minor histocompatibility antigens (MiHA)-expressing leukemic cells. Unfortunately, recognition of MiHA expressed in healthy tissues, including the liver, skin and gut, often results in morbidity and even mortality due to the development of graft-versus-host disease (GVHD). Antigen presenting cells (APC) play an essential role in the activation of alloreactive T cells and thus in the induction of GVL responses and GVHD. Using several in vivo models it has been suggested that host APC, which present endogenous MiHA to donor CD8+ T cells, are more potent inducers of alloreactive responses than donor APC. Here, we compared the ability of host APC versus donor APC to activate alloreactive T cells in vitro and to induce alloreactions in vivo. For this purpose, we isolated T cells and cultured bone marrow-derived dendritic cells (DC), the most potent APC, from B6.SJL ‘host’ mice and MHC (H-2b)-matched but MiHA-mismatched C3.SW ‘donor’ mice. Following their activation by the TLR4 ligand lipopolysaccharide for no longer than 5 hours, both host and donor DC showed a fully mature phenotype (MHCIhigh MHCIIhigh CD40+ CD80high CD86high), and comparable migratory behavior towards lymph node-specific chemokines. In primary mixed leukocyte reactions host DC were 2–3 times more potent than donor DC in inducing donor CD8+ T cell proliferation. To determine the role of host and donor DC in the development of GVL and GVHD, we established a preclinical mouse model (C3>B6) for allogeneic stem cell transplantation. Repopulation analyses demonstrated that T cells, B cells and myeloid cells from blood and spleen of transplanted mice were for more than 95% of donor origin. Transplanted mice were subsequently vaccinated intravenously with 3 weekly doses of 3 × 105 host or donor DC. DC vaccination, which did not lead to significant GVHD, was followed by an intravenous challenge with 106 GFP-expressing mouse AML (C1498) cells. While sham-vaccinated mice died quickly due to leukemia progression, vaccination with both host and donor DC resulted in diminished leukemia growth. However, only vaccination with host DC, and not donor DC, resulted in a (partial) rescue from challenge. These results suggest that host DC are the most potent inducers of GVL reactions without causing significant GVHD. We believe that these data will be instrumental in designing appropriate DC-based immunotherapies in transplanted leukemia patients.