The authors would like to take this opportunity to thank Dr. Sackstein [1Sackstein R. Re: Ex vivo fucosylation improves human cord blood engraftment in NOD-SCID IL-2Rγnull mice.Exp Hematol. 2012; 40: 518-519Abstract Full Text Full Text PDF PubMed Scopus (5) Google Scholar] for his insightful comments regarding the use of manganese (Mn2+) in our cord blood experiments involving ex vivo treatment with fucosyltransferase (FT)-VI, published in this issue of Experimental Hematology [2Robinson S.N. Simmons P.J. Thomas M.W. et al.Ex vivo fucosylation improves human cord blood engraftment in NOD-SCID IL-2Rγnull (NSG) mice.Exp Hematol. 2012; 40: 445-446Abstract Full Text Full Text PDF PubMed Scopus (64) Google Scholar].The authors are pleased that their observations on the hematotoxicity of Mn2+ are consistent with those of Sackstein et al. Given the hematotoxicity of 10 mM Mn2+, experiments were performed under significantly less hematotoxic conditions (1 mM Mn2+) with incubations performed at room temperature rather than 37°C. A more rigorous assessment of Mn2+ toxicity was considered outside the scope of the current publication. In addition, given the observation that FT-VI activity was demonstrated in the complete absence of Mn2+, it is likely that this option would be pursued in a clinical setting. No data were omitted from the homing experiment and representative dot plots of marrow samples from control mice and mice receiving buffer-treated, or FT-VI–treated CD34+CFSE+ cells were shown for comparison. The authors illustrate using bioluminescent imaging that the homing of CD34+ cells (whether treated with FT-VI, or not) occurs throughout the hematopoietic system of the mouse, supporting our hypothesis that ex vivo fucosylation of cord blood CD34+ cells does not alter the distribution of the engrafting cells, rather it increases the magnitude of the engraftment.The authors thank the Editor of Experimental Hematology for the opportunity to respond to the comments of Dr. Sackstein and hope that the points discussed provide clarification and interest for the reader. The authors would like to take this opportunity to thank Dr. Sackstein [1Sackstein R. Re: Ex vivo fucosylation improves human cord blood engraftment in NOD-SCID IL-2Rγnull mice.Exp Hematol. 2012; 40: 518-519Abstract Full Text Full Text PDF PubMed Scopus (5) Google Scholar] for his insightful comments regarding the use of manganese (Mn2+) in our cord blood experiments involving ex vivo treatment with fucosyltransferase (FT)-VI, published in this issue of Experimental Hematology [2Robinson S.N. Simmons P.J. Thomas M.W. et al.Ex vivo fucosylation improves human cord blood engraftment in NOD-SCID IL-2Rγnull (NSG) mice.Exp Hematol. 2012; 40: 445-446Abstract Full Text Full Text PDF PubMed Scopus (64) Google Scholar]. The authors are pleased that their observations on the hematotoxicity of Mn2+ are consistent with those of Sackstein et al. Given the hematotoxicity of 10 mM Mn2+, experiments were performed under significantly less hematotoxic conditions (1 mM Mn2+) with incubations performed at room temperature rather than 37°C. A more rigorous assessment of Mn2+ toxicity was considered outside the scope of the current publication. In addition, given the observation that FT-VI activity was demonstrated in the complete absence of Mn2+, it is likely that this option would be pursued in a clinical setting. No data were omitted from the homing experiment and representative dot plots of marrow samples from control mice and mice receiving buffer-treated, or FT-VI–treated CD34+CFSE+ cells were shown for comparison. The authors illustrate using bioluminescent imaging that the homing of CD34+ cells (whether treated with FT-VI, or not) occurs throughout the hematopoietic system of the mouse, supporting our hypothesis that ex vivo fucosylation of cord blood CD34+ cells does not alter the distribution of the engrafting cells, rather it increases the magnitude of the engraftment. The authors thank the Editor of Experimental Hematology for the opportunity to respond to the comments of Dr. Sackstein and hope that the points discussed provide clarification and interest for the reader. Re: “Ex vivo fucosylation improves human cord blood engraftment in NOD-SCID IL-2Rγnull mice”Experimental HematologyVol. 40Issue 7PreviewIt is well-established that expression of E-selectin on marrow microvascular endothelial cells mediates homing of hematopoietic stem cells to marrow (reviewed in [1]). In this issue of Experimental Hematology, Robinson et al. [2] present data corroborating the reports of others [3,4] showing that ex vivo cell surface glycan engineering achieved by treating human cord blood (CB) CD34+ cells with the sugar donor GDP-fucose and the enzyme fucosyltransferase VI enforces expression of sialylated Lewis X (sLex), the prototypical carbohydrate binding determinant for E-selectin. Full-Text PDF Open Archive

Though remissions have been observed following allo-HSCT for the treatment of CLL, many CLL patients are ineligible for transplant due to the lack of HLA-compatible donors. The use of umbilical cord blood (UCB) permits transplantation of many patients who lack HLA-compatible donors due to reduced requirements for stringent HLA matching between graft and recipient; however, disease relapse remains a concern with this modality. The generation of CLL-specific CTL from UCB T-cells, primed and expanded against the leukemic clone, might enhance the GVL effect and improve outcomes with UCB transplantation. Here we report the generation of functional, CLL-specific CTL using CD40-ligated CLL cells to prime partially-HLA matched UCB T-cells. Functionality and specificity were demonstrated by immune synapse assay, IFN-γ ELISpot, multi-parametric intracellular cytokine flow cytometry, and 51Cr release assay. The use of patient-specific, non-CLL controls demonstrated the generation of both alloantigen and CLL-specific responses. Subsequently, we developed a clinically-applicable procedure permitting separation of alloreactive CTL from leukemia-specific CTL. Leukemia-specific CTL were able to mediate in vivo killing of CLL in humanized mice without concurrent or subsequent development of xenoGVHD. Our results demonstrate that generation of CLL-specific effectors from UCB is feasible and practical, and the results support further exploration of this strategy as a treatment modality for CLL.

Delayed engraftment remains a major hurdle after cord blood (CB) transplantation. It may be due, at least in part, to low fucosylation of cell surface molecules important for homing to the bone marrow microenvironment. Because fucosylation of specific cell surface ligands is required before effective interaction with selectins expressed by the bone marrow microvasculature can occur, a simple 30-minute ex vivo incubation of CB hematopoietic progenitor cells with fucosyltransferase-VI and its substrate (GDP-fucose) was performed to increase levels of fucosylation. The physiologic impact of CB hematopoietic progenitor cell hypofucosylation was investigated in vivo in NOD-SCID interleukin (IL)-2R gamma(null) (NSG) mice. By isolating fucosylated and nonfucosylated CD34(+) cells from CB, we showed that only fucosylated CD34(+) cells are responsible for engraftment in NSG mice. In addition, because the proportion of CD34(+) cells that are fucosylated in CB is significantly less than in bone marrow and peripheral blood, we hypothesize that these combined observations might explain, at least in part, the delayed engraftment observed after CB transplantation. Because engraftment appears to be correlated with the fucosylation of CD34(+) cells, we hypothesized that increasing the proportion of CD34(+) cells that are fucosylated would improve CB engraftment. Ex vivo treatment with fucosyltransferase-VI significantly increases the levels of CD34(+) fucosylation and, as hypothesized, this was associated with improved engraftment. Ex vivo fucosylation did not alter the biodistribution of engrafting cells or pattern of long-term, multilineage, multi-tissue engraftment. We propose that ex vivo fucosylation will similarly improve the rate and magnitude of engraftment for CB transplant recipients in a clinical setting. (C) 2012 ISEH - Society for Hematology and Stem Cells. Published by Elsevier Inc.

Abstract Abstract 246 Natural killer (NK) cells are an innate component of immune system that can produce a graft vs. leukemia (GVL) effect after stem cell transplantation. NK cells derived from acute myeloid leukemia (AML) patients are defective in their cytolytic function against leukemic cells. In order to better understand the mechanism of this defect, we performed functional assays examining immunological synapse formation of AML patient NK cells with autologous and allogeneic primary AML cells acting as antigen-presenting cells (APCs). Confocal microscopy was used to image and score F-actin polymerization at the immunological synapse between patient NK cells and leukemic cells. Accumulation of F-actin beneath the area of the NK: APC contact site is a hallmark of NK lytic synapses and allows signaling molecules to regulate appropriate activation and effector function. AML patient derived NK cells (AML-NK cells) formed significantly fewer synapses with autologous leukemia cells than healthy donor NK cells (12% versus 30%, n = 16. p > 0.001). Moreover, AML-NK cells were defective in their ability to recruit the key receptor NKG2D and the signaling molecule phosphotyrosine to immunological synapse contact sites. Signaling through the costimulatory ligand4-1BB-L (CD137L) has been shown to activate T cells, enhance antitumor responses and has multiple immunomodulatory effects on dendritic cells and NK cells. We postulated that AML-NK cells could be activated for enhanced cytolytic activity using artificial APCs generated to express CD137L. To test this, we setup co-culture assays using AML-NK cells and artificial CD137L-APCs before subsequent examination of immunological synapse function with AML blasts. Stimulated AML-NK cells that formed cell conjugate interactions with AML blasts, showed a significant increase in formation of immunological synapses compared to unstimulated AML-NK cells. The number of AML-NK/AML blast immunological synapses increased 16 hours after stimulation and peaked at approximately 72 hours. CD137L stimulation of AML-NK cells was also associated with increased cytotoxic function against primary AML cells (n = 6, p <0.01). Furthermore, CD137L stimulation increased recruitment of tyrosine-phosphorylated proteins at AML-NK immunological synapses compared with unstimulated control experiments (RRI 4.1 versus 2.3, n = 3, p < 0.01). Taken together, our data suggests that immune functional suppression of AML-NK cells in leukemia patients can be reversed by CD137L activation signaling, resulting in enhanced F-actin synapse formation, phosphotyrosine signaling, and cytolytic function. Thus, enhanced recruitment of signaling molecules to the NKIS may represent a novel immunomodulatory function of CD137L in the NK cell–mediated killing of AML cells. These findings should aid development of new immune based therapies for leukemia. Disclosures: Gribben: Roche: Honoraria; Celgene: Honoraria; GSK: Honoraria; Mundipharma: Honoraria; Gilead: Honoraria; Pharmacyclics: Honoraria.

Abstract Abstract 1794 Immune dysfunction is a hallmark of chronic lymphocytic leukemia (CLL) including suppressed humoral and cell-mediated immune responses. The immunomodulatory agent lenalidomide has shown effective clinical activity against CLL, but its mechanism of action is poorly understood. Previous work has demonstrated that the T cell immunological synapse and functional defects in CLL can be reversed following lenalidomide treatment (J Clin Invest. 2008; 118). Polymerization of F-actin at the NK cell immunological synapse with tumor cells is required for signaling molecules to assemble and regulate NK cell activation and effector function. Confocal microscopy was used to visualize and analyze F-actin polymerization at the immune synapse between NK cells and CLL cells. The impaired immune synapse defect identified in CLL could result from not only the defects of CLL B cells but also defects in the CLL NK cells or a combination of both factors. To investigate the contribution of each factor, we examined synapse formation in experiments using CLL B cells with autologous CLL NK cells or healthy allogeneic NK cells. Conjugates formed with healthy NK cells and CLL B cells exhibited a strong band of F-actin at the immune synapse. In contrast, significantly less actin polymerization at the synapse was observed in autologous CLL NK cells and CLL B cells (P < 0.01). These results indicate CLL B cells, together with CLL NK cells contributed to the immune dysfunction in CLL. As autologous NK cell function in CLL is suppressed, we investigated the utility of CB as a potential functional source of NK cells for CLL immunotherapy. We examined the effect of lenalidomide on NK cell immune synapse function with CLL B cells acting as APCs. We demonstrated that ex vivo treatment of CLL cells with lenalidomide (500 ng/ml) for 48 hours caused a significant increase in the ability of autologous CLL NK cells to form F-actin immune synapses with CLL B cells. The same treatment of CLL B cells also significantly increased the ability of CB-NK cells to form F-actin immunological synapses with these treated CLL B cells compared to untreated CLL B cells (33.6% to 67.3%, P < 0.01, n=6). Our results also show that lenalidomide treatment of autologous NK cells from CLL patients enhanced synapse formation with treated CLL cells compared to experiments using untreated NK cells, but with reduced function compared to CB NK cells. Of note, lenalidomide treatment was shown to increase the recruitment of the signaling molecule Lck to NK cell:CLL cell synapse site, that is known to regulate lytic synapse function. Importantly, lenalidomide treatment significantly increased CB-NK killing of CLL B cells compared to untreated CLL B cells (20.5% versus 48.2%, E:T ratio of 10:1, n = 6, p < 0.001). These results provide insight into the potential mechanism of action of lenalidomide's anti-leukemic function – priming CLL tumor cells for enhanced NK cell lytic synapse formation and effector function. In addition, the data suggests that immunotherapeutic strategies utilizing a combination of CB-NK cells and lenalidomide has an enhanced clinical efficacy in CLL. Disclosures: Gribben: Roche: Honoraria; Celgene: Honoraria; GSK: Honoraria; Mundipharma: Honoraria; Gilead: Honoraria; Pharmacyclics: Honoraria.

Abstract Abstract 1370 Background: Though the cancer immune surveillance hypothesis was first proposed a century ago, there has been limited evidence to support the role of antigen presentation in the detection or suppression of CLL. In this study we evaluated the frequencies of HLA haplotype and homozygosity and subsequent impact on clinical outcome in CLL patients with advanced disease. Methods: We performed a retrospective chart review of 249 CLL patients who were referred for allogeneic stem cell transplant at MD Anderson Cancer Center. We compared HLA allele frequencies of the patient population with those of local, race-matched controls and identified specific HLA alleles which were more frequent in the patient population. We also compared HLA homozygosity between the patient and control population. The Kaplan-Meier method was then used to determine the prognostic significance of the identified HLA alleles and homozygosity on clinical outcome within our patient population. Progression-free survival (PFS) was calculated from the time of first treatment to the time of progression or death. Results: CLL patients with advanced disease were significantly more likely to express HLA-A1 (OR=1.49, 95% CI 1.15–1.94, p=0.0003) or HLA- C7 (OR 1.24, 95% CI 1.00–1.53, p=0.05). In addition, these patients were more likely to be homozygous at any HLA locus than were controls (OR=1.20, 95% CI 0.97–1.48, p=0.04), particularly at HLA-C (OR=1.62, 95% CI 1.13–2.33, p=0.002) and at multiple HLA loci (OR=1.69, 95% CI 1.06–2.70, p=0.006). CLL patients who were HLA-A1+, HLA-A1/C7+ or homozygous at any allele demonstrated worse PFS in comparison with CLL patients without any of these HLA allelic characteristics. Median survival was 23.9 months for HLA-A1+ patients, 13.9 months for HLA-A1/C7+ patients and 25.7 months for patients with homozygosity, in comparison to 31.8 months for the population without any detrimental alleles or homozygosity (p=0.02, p=0.0008, and p=0.007 respectively, Figure 1: A, B, C). Analysis of patients possessing only HLA-C7 as a risk factor demonstrated a trend toward decreased PFS but was not quite statistically significant (p=0.07, data not shown). Conclusions: Patients with advanced CLL appear to express certain HLA alleles and exhibit HLA homozygosity more frequently than normal controls. In addition, these HLA characteristics may predispose CLL patients to a worse outcome. Because HLA allelic variation determines the specificity of antigens presented to the immune system, the data suggest that immune surveillance may play a physiologic role in the control of leukemic disease and provide a theoretical framework for the identification of CLL antigens which could eventually serve as targets for immunotherapy. A. Negative effects of HLA-A1 allele on overall survival of patients with advanced CLL are B. synergistically worsened by the presence of the HLA-C7 allele. C. Homozygosity at any HLA allele also imparted a negative impact upon overall survival. Disclosures: O'Brien: Novartis: Research Funding; BMS: Research Funding.

Peripheral blood natural killer (NK) cell therapy for acute myeloid leukemia has shown promise in clinical trials after allogeneic stem cell transplantation. Cord blood (CB) is another potentially rich source of NK cells for adoptive immune therapy after stem cell transplantation. Tightly regulated receptor signaling between NK cells and susceptible tumor cells is essential for NK cell-mediated cytotoxicity. However, despite expressing normal surface activating and inhibitory NK receptors, CB-derived NK cells have poor cytolytic activity. In this study, we investigate the cellular mechanism and demonstrate that unmanipulated CB-NK cells exhibit an impaired ability to form F-actin immunologic synapses with target leukemia cells compared with peripheral blood-derived NK cells. In addition, there was reduced recruitment of the activating receptor CD2, integrin leukocyte function-associated antigen-1, and the cytolytic molecule perforin to the CB-NK synapse site. Exvivo interleukin (IL)-2 expansion of CB-NK cells enhanced lytic synapse formation including CD2 and leukocyte function-associated antigen-1 polarization and activity. Furthermore, the acquired antileukemic function of IL-2-expanded CB-NK cells was validated using a nonobese diabetic severe combined immunodeficient IL-2 receptor common γ-chain null mouse model. We believe our results provide important mechanistic insights for the potential use of IL-2-expanded CB-derived NK cells for adoptive immune therapy in leukemia.

Abstract 2453 The incompatibility between donor killer cell immunoglobulin-like receptors (KIRs) and their corresponding ligands has been reported to reduce the risk of relapse after haploidentical and human leukocyte antigen (HLA) mismatched hematopoietic stem cell transplantation in patients with acute myeloid leukemia. We tested this KIR-ligand mismatch hypothesis in the context of allogeneic cord blood NK cells as an adoptive transfer of lymphocytes treating residual chronic lymphocytic leukemia (CLL). As a model for targeting malignant B cells in CLL, we examined allogeneic cord blood NK cell function in NOD scid gamma (NSG) mice, which carry the null interleukin-2 receptor gamma chain mutation, as the mice developed leukemia. Positively selected CD56+ cord blood NK cells were expanded ex vivo with interleukin-2 for 14 days. CLL cells were established in the NSG model by infusion of CLL cells obtained from patients. The leukemia that develops in NSG mice resembles human CLL, with a proliferating CD19+CD23+CD5+ B-cell population detected in the bone marrow, spleen, lymph nodes, and peripheral blood. Subsequently, expanded cord blood NK cells (5 × 10 6 per mouse) were intravenously infused into NSG-CLL mice. The NK cells that were infused into the CLL mice were typed for HLA and KIR (four main KIRs: KIR2DL2, KIR2DL3, KIR3DL1, and KIR2DL1). The CLL patients9 samples that had been used in the NSG models were genotyped for KIR ligands (HLA-C group or HLA-Bw4 group and HLA-A3). In the six pairs of cord blood NK and CLL cells typed, all were HLA mismatched. Five pairs were KIR-ligand mismatched; these mice showed robust NK cell–mediated killing of CLL cells 7 days after NK cell infusion. Of interest, although no KIR-ligand mismatch was seen between the cord blood NK cells and CLL cells in one pair, we still observed NK cell–mediated killing of CLL cells in the mice. In this instance, NK cell–mediated cell killing could have been attributed to possible lower expression of HLA ligands by leukemic cells. Overall survival was significantly improved in CLL-NSG mice that had received cord blood NK cell treatment compared with overall survival in untreated mice (Kaplan-Meier analysis, p Disclosures: Gribben: Roche: Consultancy; Celgene: Consultancy; GSK: Honoraria; Napp: Honoraria.

Cord blood (CB) provides a readily available, ethnically-diverse source of hematopoietic tissue for transplantation, however, low cell dose delays engraftment and increases risk of graft failure. In addition, it has also been suggested that CB hematopoietic stem cells (HSC) home less efficiently to the hematopoietic microenvironment after transplantation when compared to bone marrow or peripheral blood HSC. Homing mechanisms may rely, at least in part, on the interaction between fucosylated glycoproteins on the surface of HSC and P- and E-selectins expressed by the microvasculature of the hematopoietic system. Glycoproteins expressed on the surface of CB HSC appear to be poorly fucosylated. It has been proposed that the use of a fucosyltransferase (FT) to increase levels of surface glycoprotein fucosylation may improve homing to the bone marrow and thereby improve engraftment. In an initial series of experiments, levels of fucosylation of surface glycoproteins expressed on CB CD34+ cells were increased using FT-VI (Engraftin™, America Stem Cell). When subsequently injected into an NOD-SCID IL-2Rγnull mouse model, the rate and magnitude of human engraftment was greater than in mice receiving a similar dose of untreated CB CD34+ cells. Since CB recipients receive mononuclear cells (MNC) rather than selected CD34+ cells, the efficacy of a fucosylation strategy using CB MNC was investigated. Methods: MNC were obtained from a frozen CB unit. Half was untreated and half fucosylated: incubation for 30 minutes at room temperature with 1 mM GDP β-Fucose, 1 mM MnCl2 and FT-VI. Fucosylated and untreated MNC were washed and each transplanted into sublethally-irradiated (270 cGy) NOD-SCID IL-2Rγnull mice (5 mice/gp). Mice each received 6 × 106 CB MNC (containing 2.5 × 104 CD34+ cells) by tail vein injection. Engraftment was followed at weekly intervals by flow cytometric measurement of human CD45+ cells in the peripheral blood. Results: 6 weeks after transplantation, recipients of fucosylated MNC showed a maximal 14% human engraftment, as compared to 1% in recipients of untreated MNC. Conclusion: Consistent with preliminary data following transplantation of fucosylated, or untreated CB CD34+ cells, fucosylation of CB MNC appears to improve both the rate and magnitude of engraftment. The efficacy of this technology will be explored in both the single and double CB transplant settings and a clinical trial evaluating this approach will be designed and executed.

Abstract Abstract 2370 Poster Board II-347 Adoptive transfer of NK cells in human B-CLL has been limited by the lack of suitable animal models to test the clinical efficacy of this immune therapy strategy. Primary patient B-CLL cells are difficult to engraft in NOD-SCID mice as these mice lack an immune microenvironment that provides essential accessory cells for tumor development. In the current study, we utilize a novel 2-step engraftment protocol using NOD-SCID γnull mice (NSG). Firstly, human cord blood (CB) derived CD34+ stem cells were engrafted to generate humanized chimeric mice capable of supporting B-CLL cells. By week 12, the human engraftment level reached 30% to 60% as detected in peripheral blood. Secondly, these mice were infused with human primary B-CLL cells labeled with CFSE. Our results show that following tail vein injection of 5E7 primary patient CLL cells, chimeric mice exhibited proliferation of CFSE+CD5+ B cells in the spleen and bone marrow, with disease development resembling human CLL. This allowed us to investigate the efficacy of ex vivo IL-2 expanded CB NK cell therapy using this novel mouse model system. As early as 24 hours post-infusion of IL-2 expanded human CD56+CD3- CB NK cells (10E7 per mouse) there was a detectable reduction of CD5+ leukemia cells in the peripheral circulation of CLL-engrafted mice by flow cytometry analysis. Moreover, by day 7 there was a dramatic 99.5 % reduction of CD5+ B-CLL cells in the blood, bone marrow, and spleen of experimental animals compared to the non-treated control group (P<0.001). Of note, expanded CB-NK cells exhibited high anti-leukemic specificity as healthy B cells from the original CB transplant were spared post-infusion. Overall, our studies suggest that the chimeric NSG mouse has utility as an in vivo model for testing immunotherapeutic strategies in aggressive B-CLL. Here, our results highlight the strong anti-leukemic response of infused ex vivo expanded CB-NK cells. Disclosures: No relevant conflicts of interest to declare.

In the control of T-helper type I (Th-1) polarization, dendritic cells (DCs) must interpret a complex array of stimuli, many of which are poorly understood. Here we demonstrate that Th-1 polarization is heavily influenced by DC-autonomous phenomena triggered by the loading of DCs with antigenically matched major histocompatibility complex (MHC) class I and class II determinants, that is, class I and II peptide epitopes exhibiting significant amino acid sequence overlap (such as would be physiologically present during infectious processes requiring Th-1 immunity for clearance). Data were derived from 13 independent antigenic models including whole-cell systems, single-protein systems, and 3 different pairs of overlapping class I and II binding epitopes. Once loaded with matched class I and II antigens, these "Th-1 DCs" exhibited differential cytokine secretion and surface marker expression, a distinct transcriptional signature, and acquired the ability to enhance generation of CD8(+) T lymphocytes. Mechanistically, tRNA-synthetases were implicated as components of a putative sensor complex involved in the comparison of class I and II epitopes. These data provide rigorous conceptual explanations for the process of Th-1 polarization and the antigenic specificity of cognate T-cell help, enhance the understanding of Th-1 responses, and should contribute to the formulation of more effective vaccination strategies.

Abstract Abstract 2663 Poster Board II-639 Natural killer (NK) cells are an important component of the innate immune surveillance of tumor cells. Defective NK cell function has been correlated with poor prognosis in acute myeloid leukemia (AML). It is well established that NK cell-mediated cytolytic activity is significantly diminished in AML patients; the mechanisms of this hypo-function are not well understood. Identifying mechanisms of tumor-induced immune suppression of lymphocytes function will aid the development of effective immunotherapeutic strategies. In the present study we examined the molecular basis for impaired NK cell responses in AML and demonstrate impaired NK cell immunological synapse formation. Confocal microscopy was used to visualize F-actin polymerization at the immune synapse between CD56+ CD3- NK cells and autologous AML blasts. We identified a significant reduction in formation of the NK cell immune synapse (NKIS) (p<0.001) from AML patients compared healthy donors (> 70% reduction). This defect was induced by direct tumor contact since NK cell defects were induced in healthy NK cells when they were co-cultured (in direct contact) for 48 hr with allogeneic AML blasts, but not with healthy allogeneic monocytes (P < 0.01). In control transwell co-culture experiments, where the NK cells and AML blast were not in direct contact, we did not observe the induced defect. We examined the molecular nature of the AML blast induced defect by quantifying recruitment of a number of these NK cell adhesion and cytoskeletal signaling proteins to the immune synapse. Following primary co-culture with AML blasts, healthy NK cells showed significantly reduced recruitment of integrin LFA-1, CD2, Lck, WASP, and tyrosine-phosphorylated protein to the NK-AML target interactions synapse (P < 0.001). These studies demonstrate a role for the tumor induced immune suppression of NK cells and will aid in the development of immunotherapeutic approaches targeting AML. Disclosures: No relevant conflicts of interest to declare.

Activated natural killer (NK) cells derived from cord blood (CB) have been reported to mediate a significant graft versus leukemia effect (GVL). However unmanipulated CB NK cells exhibit poor cytolytic activity against tumor cells in vitro, thus limiting their clinical application. We investigated the mechanism for the poor cytolytic activity of CB NK cells and whether the defect can be overcome with ex vivo expansion. NK cell killing of the tumor target cells is achieved by the formation of a mature immune synapse, followed by secretion of lytic granules containing perforin and granzymes. We hypothesized that CB NK cells exhibit low cytotoxicity against leukemia blasts due to a defect in the formation of the immune synapse. We have found reduced ability of CB NK cells to form immune synapses with leukemia target cells contributing to the decreased cytotoxicity of unmanipulated CB NK cells compared to adult peripheral blood (APB) NK cells. F-actin polarization was observed in a mean of 12% (range 9–21%) of the CB NK cell/tumor cell conjugates versus a mean of 85% (range 68–87) of APB NK/tumor conjugates (p<0.001). This impairment could then be reversed by ex vivo expansion of CB NK cells with IL-2. IL-2-expanded CB NK cells formed increased immune synapses with K562 tumor cells (mean 65%; 60–71%) and primary human AML blasts (mean 48%, range 39–55%), comparable to the levels generated with peripheral blood NK cells. Additionally, we demonstrated that ex vivo expanded CB NK cells could efficiently kill human AML engrafted in a NOD/scid/IL-2Rg-null mouse model. A mean 50% (37–66%) reduction in AML blasts was observed in comparison to control groups (mean 3%;range 2.4–4.7%) by 6 weeks post NK infusion (p<0.05). Our results suggest that ex vivo expansion of CB NK cells is a feasible and effective strategy for the treatment of AML.

The use of umbilical cord blood (UCB) grafts for hematopoietic stem cell transplantation (HSCT) is a promising technique that permits a degree of human leukocyte antigen mismatch between the graft and the host without the concomitant higher rate of graft-versus-host disease that would be observed between an adult marrow graft and a mismatched host. A disadvantage to the use of UCB for HSCT is that immune reconstitution may be significantly delayed because of the low stem cell dose available in the graft. Ex vivo expansion of UCB CD34 cells would provide a greater number of stem cells; however, there are persistent concerns that ex vivo-expanded CD34 cells may lose pluripotency and the ability to contribute meaningfully to long-term engraftment. To address this issue, we transduced CD34-selected UCB cells with a lentiviral construct expressing luciferase, and determined homing and engraftment patterns in vivo by noninvasive bioluminescent imaging in sublethally irradiated NOD/SCID/IL-2R gamma(-/-) (NSG) mice. Graft contribution to multilineage commitment was also confirmed by analysis of primary and secondary transplants by flow cytometry and immunohistochemistry. Our results demonstrate that, other than a mild delay at the onset of engraftment, there were no significant differences in lineage repopulation or in long-term or secondary engraftment between culture-expanded and unexpanded UCB CD34-selected cells. The results suggest that multipotent stem cells can be expanded ex vivo and can contribute meaningfully to long-term hematopoietic engraftment. STEM CELLS 2009;27:1932-1940

Introduction: Long-term remissions have been observed following allogeneic hematopoietic stem cell transplantation (allo-HSCT) for the treatment of chronic lymphocytic leukemia (CLL). Unfortunately, many CLL patients are ineligible for transplant due to the lack of an HLA-compatible donor. Umbilical cord blood (UCB) HSCT permits transplantation of many individuals who are ineligible for allo-HSCT due to its reduced requirement for stringent HLA matching; however, disease relapse remains a significant complication. Further, unlike allo-HSCT, donor lymphocyte infusion (DLI) cannot be employed as a post-transplant therapy. To address this issue we are developing a strategy to generate CLL-specific T-lymphocytes to be employed as a post-UCB-HSCT therapeutic treatment. Methods: CLL-specific antigen presenting cells (APC) were generated by the transduction of CLL cells with an adenoviral vector encoding CD154 (CD40L). Successful transduction and subsequent CD40 ligation in CLL cells were verified by monitoring CD95 expression. CLL APC were then used to prime partially HLA matched (typically 4/6) UCB lymphocytes in the presence of IL-12, IL-2, IL-7, and IL-15. Expanded UCB lymphocytes were phenotyped by flow cytometry, and CLL specificity was determined by ELISpot and 51Cr lysis using non-CLL (CD19-/CD3+) patient lymphocytes as allo-antigen controls. Results: Effector cell phenotype was predominantly (>80%) CD4+; however, skewing toward CD8+ expansion could be achieved by incubation in IL-12 during priming. Significant populations of effector memory (CD62L+CD127+) and central memory (CD62L-CD127+) cells were observed following repeated stimulations with CLL-APC. Significant expansion (>4 fold) of UCB lymphocyte populations was typically observed. In ELISpot assays, CLL-APC primed UCB lymphocyte responders exhibited a significant increase (p > 0.05) in IFN-γ ELISpots when incubated with CLL stimulators in comparison to non-CLL control stimulators. Cytolytic activity was demonstrated against unmodified CLL cells by 51Cr release assays in which percent lysis of CLL targets was 30–50% at E:T ratios of 10:1 and 20:1. Lysis of autologous non-CLL control targets was negligible. Conclusions: Our results demonstrate that in vitro generation of CLL-specific effectors from partially HLA-matched UCB lymphocytes is both feasible and practical. These preclinical results support further exploration of this technique as a promising treatment modality in conjunction with UCB-HSCT.

Cord blood (CB) is a viable alternative to bone marrow (BM) or mobilized peripheral blood (mPB) for transplantation. While CB is collected, typed, frozen and banked as a readily available source of ethnically-diverse tissue for transplant, a major limitation of CB transplantation is low cell dose leading to delayed engraftment and increased risk of engraftment failure. Ex vivo expansion of CB could improve transplant outcomes by increasing cell dose. Using a mesenchymal stem cell (MSC) co-culture system, significant levels of CD34+ cell expansion can be achieved. However, the presence of non-CD34+ accessory cells may impact the efficiency of expansion. Thawed CB units contain (mean ± SEM, n = 114) 1.4 ± 0.6% CD34+, 39.5 ± 1.2% CD3+, 17.1 ± 0.7% CD19+, 15.3 ± 0.7% CD14+ and 20.1 ± 1.2% CD56+ cells. Preliminary experiments demonstrated that only CD3+ and/or CD14+ inhibited CD34+ expansion. No positive or negative impact was observed for any other cell population. These data suggested CD3+ and CD14+ cells as candidate populations for investigation. Our hypothesis is that the depletion of CD3+ and/or CD14+ cells from CB MNC will markedly improve CD34+ expansion in the MSC co-culture system.Methods: CD3+ and/or CD14+ cells were removed from a single CB unit by magnetic depletion (MACS, Miltenyi) and the products cultured in the MSC co-culture system. Comparisons were made between the CD34+ and total nucleated cell (TNC) expansion achieved in the MSC co-culture system following the incubation of MNC (nothing depleted), MNC depleted of CD3+ cells (MNC-CD3+), CB MNC depleted of CD14+ cells (MNC-CD14+) and CB MNC depleted of both CD3+ and CD14+ cells (MNC-(CD3+ + CD14+)). Incubations were performed in medium containing Flt-3L, SCF, G-CSF and TPO for 7 days. On day 7, non-adherent (n/a) cells were removed and cultured in liquid culture conditions (no MSC) for a further 7 days. On Day 14, n/a cells from the co-culture flask and liquid cultures were pooled and evaluated for TNC and CD34+ cell content.Tabled 1Fold increase over input (range: min-max, n=3)TNCCD34+MNC3.5-19.811.3-12.4MNC-CD3+21.2-47.419.2-51.3MNC-CD14+8.3-27.023.6-32.4MNC-(CD3+ + CD14+)27.8-41.023.9-55.1 Open table in a new tab Conclusion: The depletion of CD3+ and/or CD14+ cells from the CB MNC increased ex vivo expansion of TNC and CD34+ when compared with original CB MNC expansion. GMP-compliant antibodies that target CD3 and CD14 are available (Miltenyi) and could be used for clinical evaluation of this strategy. Cord blood (CB) is a viable alternative to bone marrow (BM) or mobilized peripheral blood (mPB) for transplantation. While CB is collected, typed, frozen and banked as a readily available source of ethnically-diverse tissue for transplant, a major limitation of CB transplantation is low cell dose leading to delayed engraftment and increased risk of engraftment failure. Ex vivo expansion of CB could improve transplant outcomes by increasing cell dose. Using a mesenchymal stem cell (MSC) co-culture system, significant levels of CD34+ cell expansion can be achieved. However, the presence of non-CD34+ accessory cells may impact the efficiency of expansion. Thawed CB units contain (mean ± SEM, n = 114) 1.4 ± 0.6% CD34+, 39.5 ± 1.2% CD3+, 17.1 ± 0.7% CD19+, 15.3 ± 0.7% CD14+ and 20.1 ± 1.2% CD56+ cells. Preliminary experiments demonstrated that only CD3+ and/or CD14+ inhibited CD34+ expansion. No positive or negative impact was observed for any other cell population. These data suggested CD3+ and CD14+ cells as candidate populations for investigation. Our hypothesis is that the depletion of CD3+ and/or CD14+ cells from CB MNC will markedly improve CD34+ expansion in the MSC co-culture system. Methods: CD3+ and/or CD14+ cells were removed from a single CB unit by magnetic depletion (MACS, Miltenyi) and the products cultured in the MSC co-culture system. Comparisons were made between the CD34+ and total nucleated cell (TNC) expansion achieved in the MSC co-culture system following the incubation of MNC (nothing depleted), MNC depleted of CD3+ cells (MNC-CD3+), CB MNC depleted of CD14+ cells (MNC-CD14+) and CB MNC depleted of both CD3+ and CD14+ cells (MNC-(CD3+ + CD14+)). Incubations were performed in medium containing Flt-3L, SCF, G-CSF and TPO for 7 days. On day 7, non-adherent (n/a) cells were removed and cultured in liquid culture conditions (no MSC) for a further 7 days. On Day 14, n/a cells from the co-culture flask and liquid cultures were pooled and evaluated for TNC and CD34+ cell content. Conclusion: The depletion of CD3+ and/or CD14+ cells from the CB MNC increased ex vivo expansion of TNC and CD34+ when compared with original CB MNC expansion. GMP-compliant antibodies that target CD3 and CD14 are available (Miltenyi) and could be used for clinical evaluation of this strategy.

Abstract Abstract 3029 Poster Board II-1005 Donor peripheral blood (PB) natural killer (NK) cell have shown clinical promise in cancer immunotherapy. Tightly regulated receptor signaling between NK cells and susceptible tumor cells is essential for NK cell-mediated cytotoxicity. Umbilical cord blood (CB) represents an important alternative source of NK cells for adoptive immune therapy. We first demonstrated that cord blood (CB) derived NK cells have poor cytolytic activity and deficiency in the formation of the F-actin immunological synapse with HLA class I deficient target K562 cells and primary AML blasts compared to PB-NK cells. In this study, we explored the cellular mechanism of these dysfunctions. We hypothesized that adhesion and signaling molecules may be defective in unmanipulated CB NK cells. Activating receptor Both CD2 and the integrin lymphocyte function-associated antigen (LFA-1) play important roles in both T lymphocyte and NK cell immune synapse formation and their trafficking to the immune synapse regulates both T and NK cell function. We now show that unmanipulated CB NK cells exhibit reduced LFA-1 mediated adhesion to mobilized ICAM-1 compared to IL-2 expanded CB NK cells (CB NK 29.7+/- 3.2 %, vs expanded CB NK 78.5+/- 6.1%, n=6). Moreover, unmanipulated CB-NK cells demonstrated reduced surface expression of CD2, and high affintyLFA-1 detected by the specific antibody (MHM24). There was decreased recruitment of CD2 and LFA-1 to the NK cell immune synapse site as quantified by confocal microscope analysis (RRI CD2 CB NK 2.02 vs PB NK 4.98, n=3). Furthermore, defective LFA-1 trafficking lead to a decrease in downstream cytotoxic granules that traffic to the immunological synapse as demonstrated by decreased perforin trafficking to the CB-NK synapse site (> 60% reduction).We next wanted to confirm that CD2 or LFA-1 play a role in restoring the immune synapseformation for IL-2 expanded CB NK cells. We incubated expanded CB NK cells with blocking antibodies specific for LFA-1 or CD2 prior to conjugation to the K562 target cells. After CD2 or LFA-1 blocking there was decreased synapse formation, with a resultant decrease in cytotoxic function. When monoclonal antibodies against both CD2 and LFA-1 were used there was significant blockade of the formation of the immune synapse, and a marked reduction of CB NK cell cytolytic activity (Mean specific lysis of K562 targets at E:T ratio 20:1 was 81% IgG control vs 22% with anti-CD2; and 29% with anti-LFA-1, n=6, P<0.001). This data shows that CD2 and LFA-1 are defective in unmanipulated CB NK cells resulting in impaired immune synapse formation. In contrast, ex vivo IL-2 expansion of CB-NK cells enhanced lytic synapse formation with the synergistic repair of CD2 and LFA-1 localization and activity. We believe our results provide important mechanistic insights for the potential use of IL-2 expanded CB-derived NK cells for adoptive immune therapy in leukemia. Disclosures No relevant conflicts of interest to declare.

Cord blood (CB) is used increasingly in transplant patients lacking sibling or unrelated donors. A major hurdle in the use of CB is its low cell dose, which is largely responsible for an elevated risk of graft failure and a significantly delayed neutrophil and platelet engraftment. As a positive correlation has been shown between the total nucleated cell (TNC) and CD34(+) cell dose transplanted and time to neutrophil and platelet engraftment, strategies to increase these measures are under development. One strategy includes the ex vivo expansion of CB mononuclear cells (MNC) with MSC in a cytokine cocktail. We show that this strategy can be further improved if CD3(+) and/or CD14(+) cells are first depleted from the CB MNC before ex vivo expansion. Ready translation of this depletion strategy to improve ex vivo CB expansion in the clinic is feasible as clinical-grade devices and reagents are available. Ultimately, the aim of improving TNC and CD34(+) transplant doses is to further improve the rate of neutrophil and platelet engraftment in CB recipients.

Cancer patients and recipients of hematopoietic stem cell transplantation exhibit a negligible response to influenza vaccine. Toward the goal of addressing this issue, we developed an in vitro model of dendritic cell (DC) immunotherapy utilizing DCs generated from naïve umbilical cord blood (UCB). UCB DCs were loaded with purified rHA protein and used to stimulate autologous T-lymphocytes. Upon recall with HA-loaded autologous DC, a 4–10-fold increase in the number of IFN-γ producing T-lymphocytes was observed in comparison to T-cells stimulated with control DCs. Antigen-specific T-cell functionality was determined by 51Cr lytic assay. Using a peptide library of predicted HA binding epitopes, we mapped an HA-specific, DR15-restricted CD4 T-cell epitope and observed tetramer positive cells. This model demonstrates that HA-specific immune responses might possibly be generated in a de novo fashion and suggests that dendritic cell immunotherapy for the prevention of influenza in populations of immunosuppressed individuals could be feasible.