Search

IL-15 is a proinflammatory cytokine that plays an essential role in the development and activation of natural killer (NK) cells. Adipose tissue acts as an endocrine organ that secretes cytokines and is an important reservoir for lymphocytes. We hypothesized that activation of the IL-15 signaling in adipose tissue will activate and expand the NK cell population and control tumor growth. We recently developed an adipocyte-targeting recombinant adeno-associated viral (rAAV) vector with minimal off-target transgene expression in the liver. Here, we used this rAAV system to deliver an IL-15/IL-15Rα complex to the abdominal fat by intraperitoneal (i.p.) injection. Adipose IL-15/IL-15Rα complex gene transfer led to the expansion of NK cells in the adipose tissue and spleen in normal mice without notable side effects. The i.p. injection of rAAV-IL-15/IL-15Rα complex significantly suppressed the growth of Lewis lung carcinoma implanted subcutaneously and exerted a significant survival advantage in a B16-F10 melanoma metastasis model. The antitumor effects were associated with the expansion of the NK cells in the blood, spleen, abdominal fat, and tumor, as well as the enhancement of NK cell maturity. Our proof-of-concept preclinical studies demonstrate the safety and efficacy of the adipocyte-specific IL-15/IL-15Rα complex vector as a novel cancer immune gene therapy.

MicroRNAs (miRNAs) bind to complementary sequences of target mRNAs, resulting in translational repression or target degradation and thus gene silencing. miRNAs are abundant in circulating blood, yet it is not known whether, as a class of regulatory molecules, they interact with human natural killer (NK) cells. Here we found that the treatment of human NK cells with several mature miRNAs in the presence of a low concentration of interleukin-12 induced CD69 expression, interferon-γ production, and degranulation marker CD107a expression. In vivo, infusion of several miRNAs alone in murine peripheral blood also resulted in comparable NK-cell activation, but not T-cell activation. Furthermore, miRNA administration significantly protected mice from tumor development in an NK cell–dependent manner. Mechanistically, we found that miRNA stimulation led to downstream activation of nuclear factor κB (NF-κB), an effect that was blunted by a block in Toll-like receptor 1(TLR1) signaling and attenuated in lymphoma patients. Knockdown of TLR1 resulted in less activation by miRNAs. Collectively, we show that miRNAs have a capacity to selectively activate innate immune effector cells that is, at least in part, via the TLR1–NF-κB signaling pathway. This may be important in the normal host defense against infection and/or malignant transformation.

MicroRNAs (miRNAs) bind to complementary sequences of target mRNAs, resulting in translational repression or target degradation and thus gene silencing. miRNAs are abundant in circulating blood, yet it is not known whether, as a class of regulatory molecules, they interact with human natural killer (NK) cells. Here we found that the treatment of human NK cells with several mature miRNAs in the presence of a low concentration of interleukin-12 induced CD69 expression, interferon-γ production, and degranulation marker CD107a expression. In vivo, infusion of several miRNAs alone in murine peripheral blood also resulted in comparable NK-cell activation, but not T-cell activation. Furthermore, miRNA administration significantly protected mice from tumor development in an NK cell–dependent manner. Mechanistically, we found that miRNA stimulation led to downstream activation of nuclear factor κB (NF-κB), an effect that was blunted by a block in Toll-like receptor 1(TLR1) signaling and attenuated in lymphoma patients. Knockdown of TLR1 resulted in less activation by miRNAs. Collectively, we show that miRNAs have a capacity to selectively activate innate immune effector cells that is, at least in part, via the TLR1–NF-κB signaling pathway. This may be important in the normal host defense against infection and/or malignant transformation.

The ability of natural killer (NK) cells to kill malignant or infected cells depends on the integration of signals from different families of cell surface receptors, including cytokine receptors. How such signals then regulate NK-cell cytotoxicity is incompletely understood. Here we analyzed an endogenous inhibitor of protein phosphatase 2A (PP2A) activity called SET, and its role in regulating human NK-cell cytotoxicity and its mechanism of action in human NK cells. RNAi-mediated suppression of SET down-modulates NK-cell cytotoxicity, whereas ectopic overexpression of SET enhances cytotoxicity. SET knockdown inhibits both mRNA and protein granzyme B expression, as well as perforin expression, whereas SET overexpression enhances granzyme B expression. Treatment of NK cells with the PP2A activator 1,9-dideoxy-forskolin also inhibits both granzyme B expression and cytotoxicity. In addition, pretreatment with the PP2A inhibitor okadaic acid rescues declining granzyme B mRNA levels in SET knockdown cells. Down-modulation of SET expression or activation of PP2A also decreases human NK-cell antibody-dependent cellular cytotoxicity. Finally, the induction of granzyme B gene expression by interleukin-2 and interleukin-15 is inhibited by SET knockdown. These data provide evidence that granzyme B gene expression and therefore human NK-cell cytotoxicity can be regulated by the PP2A-SET interplay.

The ability of natural killer (NK) cells to kill malignant or infected cells depends on the integration of signals from different families of cell surface receptors, including cytokine receptors. How such signals then regulate NK-cell cytotoxicity is incompletely understood. Here we analyzed an endogenous inhibitor of protein phosphatase 2A (PP2A) activity called SET, and its role in regulating human NK-cell cytotoxicity and its mechanism of action in human NK cells. RNAi-mediated suppression of SET down-modulates NK-cell cytotoxicity, whereas ectopic overexpression of SET enhances cytotoxicity. SET knockdown inhibits both mRNA and protein granzyme B expression, as well as perforin expression, whereas SET overexpression enhances granzyme B expression. Treatment of NK cells with the PP2A activator 1,9-dideoxy-forskolin also inhibits both granzyme B expression and cytotoxicity. In addition, pretreatment with the PP2A inhibitor okadaic acid rescues declining granzyme B mRNA levels in SET knockdown cells. Down-modulation of SET expression or activation of PP2A also decreases human NK-cell antibody-dependent cellular cytotoxicity. Finally, the induction of granzyme B gene expression by interleukin-2 and interleukin-15 is inhibited by SET knockdown. These data provide evidence that granzyme B gene expression and therefore human NK-cell cytotoxicity can be regulated by the PP2A-SET interplay.

Human CD56bright natural killer (NK) cells possess little or no killer immunoglobulin-like receptors (KIRs), high interferon-γ (IFN-γ) production, but little cytotoxicity. CD56dim NK cells have high KIR expression, produce little IFN-γ, yet display high cytotoxicity. We hypothesized that, if human NK maturation progresses from a CD56bright to a CD56dim phenotype, an intermediary NK cell must exist, which demonstrates more functional overlap than these 2 subsets, and we used CD94 expression to test our hypothesis. CD94highCD56dim NK cells express CD62L, CD2, and KIR at levels between CD56bright and CD94lowCD56dim NK cells. CD94highCD56dim NK cells produce less monokine-induced IFN-γ than CD56bright NK cells but much more than CD94lowCD56dim NK cells because of differential interleukin-12–mediated STAT4 phosphorylation. CD94highCD56dim NK cells possess a higher level of granzyme B and perforin expression and CD94-mediated redirected killing than CD56bright NK cells but lower than CD94lowCD56dim NK cells. Collectively, our data suggest that the density of CD94 surface expression on CD56dim NK cells identifies a functional and likely developmental intermediary between CD56bright and CD94lowCD56dim NK cells. This supports the notion that, in vivo, human CD56bright NK cells progress through a continuum of differentiation that ends with a CD94lowCD56dim phenotype.

Aberrant methylation of tumor suppressor genes can lead to their silencing in many cancers. TSC-22is a gene silenced in several solid tumors, but its function and the mechanism(s) responsible for its silencing are largely unknown. Here we demonstrate that the TSC-22promoter is methylated in primary mouse T or natural killer (NK) large granular lymphocyte (LGL) leukemia and this is associated with down-regulation or silencing of TSC-22expression. The TSC-22deregulation was reversed in vivo by a 5-aza-2′-deoxycytidine therapy of T or NK LGL leukemia, which significantly increased survival of the mice bearing this disease. Ectopic expression of TSC-22 in mouse leukemia or lymphoma cell lines resulted in delayed in vivo tumor formation. Targeted disruption of TSC-22in wild-type mice enhanced proliferation and in vivo repopulation efficiency of hematopoietic precursor cells (HPCs). Collectively, our data suggest that TSC-22normally contributes to the regulation of HPC function and is a putative tumor suppressor gene that is hypermethylated and silenced in T or NK LGL leukemia.

Aberrant methylation of tumor suppressor genes can lead to their silencing in many cancers. TSC-22 is a gene silenced in several solid tumors, but its function and the mechanism(s) responsible for its silencing are largely unknown. Here we demonstrate that the TSC-22 promoter is methylated in primary mouse T or natural killer (NK) large granular lymphocyte (LGL) leukemia and this is associated with down-regulation or silencing of TSC-22 expression. The TSC-22 deregulation was reversed in vivo by a 5-aza-2′-deoxycytidine therapy of T or NK LGL leukemia, which significantly increased survival of the mice bearing this disease. Ectopic expression of TSC-22 in mouse leukemia or lymphoma cell lines resulted in delayed in vivo tumor formation. Targeted disruption of TSC-22 in wild-type mice enhanced proliferation and in vivo repopulation efficiency of hematopoietic precursor cells (HPCs). Collectively, our data suggest that TSC-22 normally contributes to the regulation of HPC function and is a putative tumor suppressor gene that is hypermethylated and silenced in T or NK LGL leukemia.

Natural killer (NK) cells contribute to host immunity, including tumor surveillance, through the production of interferon gamma (IFN-γ). Although there is some knowledge about molecular mechanisms that induce IFN-γ in NK cells, considerably less is known about the mechanisms that reduce its expression. Here, we investigate the role of the Hlx transcription factor in IFN-γ production by NK cells. Hlx expression is induced in monokine-activated NK cells, but with delayed kinetics compared to IFN-γ. Ectopic Hlx expression decreases IFN-γ synthesis in primary human NK cells and IFN-γ promoter activity in an NK-like cell line. Hlx protein levels inversely correlate with those of STAT4, a requisite factor for optimal IFN-γ transcription. Mechanistically, we provide evidence indicating that Hlx overexpression accelerates dephosphorylation and proteasome-dependent degradation of the active Y693-phosphorylated form of STAT4. Thus, Hlx expression in activated NK cells temporally controls and limits the monokine-induced production of IFN-γ, in part through the targeted depletion of STAT4.

Natural killer (NK) cells contribute to host immunity, including tumor surveillance, through the production of interferon gamma (IFN-γ). Although there is some knowledge about molecular mechanisms that induce IFN-γ in NK cells, considerably less is known about the mechanisms that reduce its expression. Here, we investigate the role of the Hlx transcription factor in IFN-γ production by NK cells. Hlx expression is induced in monokine-activated NK cells, but with delayed kinetics compared to IFN-γ. Ectopic Hlx expression decreases IFN-γ synthesis in primary human NK cells and IFN-γ promoter activity in an NK-like cell line. Hlx protein levels inversely correlate with those of STAT4, a requisite factor for optimal IFN-γ transcription. Mechanistically, we provide evidence indicating that Hlx overexpression accelerates dephosphorylation and proteasome-dependent degradation of the active Y693-phosphorylated form of STAT4. Thus, Hlx expression in activated NK cells temporally controls and limits the monokine-induced production of IFN-γ, in part through the targeted depletion of STAT4.

Monocyte cytokines (ie, monokines) induce natural killer (NK) cells to produce interferon-γ (IFN-γ), which is critical for monocyte clearance of infectious pathogens and tumor surveillance. Human CD56bright NK cells produce far more IFN-γ in response to monokines than do CD56dim NK cells. The kinases and phosphatases involved in regulating IFN-γ production by monokine-activated NK cells are not clearly identified. SHIP1 is a 5′ inositol phosphatase that dephosphorylates the phosphatidylinositol-3 kinase (PI-3K) product PI3,4,5P3. Here, we show that constitutive expression of SHIP1 is distinctly lower in CD56bright NK cells compared with CD56dim NK cells, suggesting it could be an important negative regulator of IFN-γ production in monokine-activated NK cells. Indeed, overexpression of SHIP1 in CD56bright NK cells followed by monokine activation substantially lowered IFN-γ production. This effect was not seen when NK cells were infected with a SHIP1 mutant containing an inactive catalytic domain. Finally, NK cells in SHIP1–/– mice produced more IFN-γ in response to monokines in vivo than did NK cells from wild-type mice. Collectively, these results demonstrate that SHIP1 negatively regulates monokine-induced NK cell IFN-γ production in vitro and in vivo and provide the first molecular explanation for an important functional distinction observed between CD56bright and CD56dim human NK subsets.

Monocyte cytokines (ie, monokines) induce natural killer (NK) cells to produce interferon-γ (IFN-γ), which is critical for monocyte clearance of infectious pathogens and tumor surveillance. Human CD56brightNK cells produce far more IFN-γ in response to monokines than do CD56dimNK cells. The kinases and phosphatases involved in regulating IFN-γ production by monokine-activated NK cells are not clearly identified. SHIP1 is a 5′ inositol phosphatase that dephosphorylates the phosphatidylinositol-3 kinase (PI-3K) product PI3,4,5P3. Here, we show that constitutive expression of SHIP1 is distinctly lower in CD56brightNK cells compared with CD56dimNK cells, suggesting it could be an important negative regulator of IFN-γ production in monokine-activated NK cells. Indeed, overexpression of SHIP1 in CD56brightNK cells followed by monokine activation substantially lowered IFN-γ production. This effect was not seen when NK cells were infected with a SHIP1 mutant containing an inactive catalytic domain. Finally, NK cells in SHIP1–/–mice produced more IFN-γ in response to monokines in vivo than did NK cells from wild-type mice. Collectively, these results demonstrate that SHIP1 negatively regulates monokine-induced NK cell IFN-γ production in vitro and in vivo and provide the first molecular explanation for an important functional distinction observed between CD56brightand CD56dimhuman NK subsets.

Considerable functional heterogeneity within human natural killer (NK) cells has been revealed through the characterization of distinct NK-cell subsets. Accordingly, a small subset of CD56+NKp44+NK cells, termed NK-22 cells, was recently described within secondary lymphoid tissue (SLT) as IL-22− when resting, with a minor fraction of this population becoming IL-22+ when activated. Here we discover that the vast majority of stage 3 immature NK (iNK) cells in SLT constitutively and selectively express IL-22, a TH17 cytokine important for mucosal immunity, whereas earlier and later stages of NK developmental intermediates do not express IL-22. These iNK cells have a surface phenotype of CD34−CD117+CD161+CD94−, largely lack expression of NKp44 and CD56, and do not produce IFN-γ or possess cytolytic activity. In summary, stage 3 iNK cells are highly enriched for IL-22 and IL-26 messenger RNA, and IL-22 protein production, but do not express IL-17A or IL-17F.

Considerable functional heterogeneity within human natural killer (NK) cells has been revealed through the characterization of distinct NK-cell subsets. Accordingly, a small subset of CD56+NKp44+NK cells, termed NK-22 cells, was recently described within secondary lymphoid tissue (SLT) as IL-22−when resting, with a minor fraction of this population becoming IL-22+when activated. Here we discover that the vast majority of stage 3 immature NK (iNK) cells in SLT constitutively and selectively express IL-22, a TH17 cytokine important for mucosal immunity, whereas earlier and later stages of NK developmental intermediates do not express IL-22. These iNK cells have a surface phenotype of CD34−CD117+CD161+CD94−, largely lack expression of NKp44 and CD56, and do not produce IFN-γ or possess cytolytic activity. In summary, stage 3 iNK cells are highly enriched for IL-22 and IL-26 messenger RNA, and IL-22 protein production, but do not express IL-17A or IL-17F.

Cutaneous T-cell lymphoma (CTCL) is characterized by the expansion of malignant CD4+ T cells in the skin. There are two main subtypes of CTCL, mycosis fungoides (MF) and Sézary syndrome (SS). Previous studies have demonstrated defects in cell-mediated immunity, including altered cytokine profiles and decreased neutrophil function, and patients often have recurring bacterial and viral infections. Recent studies have shown increased expression of interleukin (IL)-15 in malignant CD4+ T cells in CTCL patients (Mishra et al., Cancer Discovery, 2016). Since IL-15 can enhance NK cell differentiation and activation, we hypothesized that NK cells from CTCL patients might have phenotypic and functional alterations due to these changes in homeostatic IL-15 levels.

Caligiuri: Innate Pharma: Membership on an entity’s Board of Directors or advisory committees. Hofmeister:Celgene Corporation: Membership on an entity’s Board of Directors or advisory committees, Speakers Bureau.

Multiple myeloma (MM) is a B-cell malignancy characterized by the aberrant clonal expansion of plasma cells (PCs) within the bone marrow (BM). Despite the use of proteasome inhibitors and immune-modulating drugs, which have improved overall survival, MM remains an incurable malignancy for which novel therapeutic approaches are urgently needed. Immunotherapy that specifically targets antigens expressed by MM would be a promising approach to treat MM patients refractory to any current treatments. Chimeric antigen receptors (CARs) are engineered fusion proteins containing tumor antigen-recognition moieties and immune cell activation domains. CAR-expressing T cells have been demonstrated successful in the clinic to treat chronic lymphocytic leukemia (CLL) and acute lymphoid leukemia (ALL). However, the potential utility of antigen-specific CAR-engineered natural killer (NK) cells or T cells to target MM-expressed CS1 to treat MM has not been previously explored, and is the focus of our study.

MicroRNAs (miRNAs) are short ribonucleic acids, which consist of an average of 22 nucleotides, and bind to complementary sequences of target mRNAs to result in translational repression or target degradation, thus silencing gene expression. MiRNAs can be abundantly found in circulating blood, yet whether, as a class of regulatory molecules, they may interact with innate immune natural killer (NK) cells has not been explored.Human NK cells were first enriched from the peripheral blood of healthy donors by negative selection using RosetteSep NK cell enrichment cocktail, followed by positive selection using anti-CD56 microbeads. After purity of ≥ 99% was confirmed by flow cytometry, NK cells were used for experiments. After being isolated from healthy donor serum by ExoQuick Exosome precipitation and verified by immunoblotting for CD9 expression, exosomes were assessed for miRNA content via real-time reverse-transcriptase (RT)-PCR using TaqMan miRNA assays. Purified NK cells were stimulated with either whole exosomes or miRNAs complexed with DOTAP, a liposomal transfection reagent. Downstream activation of Toll-like receptor (TLR) signaling by miRNAs was measured via immunoblotting for NF-kB, and its inhibition was similarly assayed in the presence of TLR blocking antibodies. Flow cytometry was used to assess NK cell activation (via CD69 surface expression) and NK cytotoxicity against tumor cells (in a CD107a degranulation assay). IFN-g production was measured via Real-time RT-PCR and enzyme-linked immunosorbent assay (ELISA). For in vivo stimulation, a complex consisting of miRNAs and Lipofectamine 2000 was administered by tail-vein injection. NK cell activation was then measured using the aforementioned in vitro assays. After in vivo stimulation with miRNAs, which was performed in the presence of NK cells or following NK depletion by TM-β1 (IL-2/15Rβ) mAb, development of implanted lymphoma tumor cells was monitored by bioluminescent imaging. NK cells purified from lymphoma patients and from healthy donors were assessed for expression of the NF-kB signaling component, p65, and TLRs via real-time RT-PCR. NK cell maturation was analyzed by flow cytometric staining for surface receptors, such as CD56 and CD94, indicative of NK cell maturation.We found that, in the presence of a low dose IL-12, treatment of human NK cells with several mature miRNAs induced CD69 expression, IFN-g production, and expression of the degranulation marker, CD107a. MiRNA-containing exosomes freshly isolated from normal human donors were also able to activate NK cells, even in the absence of IL-12. In vivo, infusion of several miRNAs into the peripheral blood similarly activated murine NK cells, while T cells were not activated. Furthermore, miRNA administration significantly protected mice from developing tumors, and this occurred in an NK cell-dependent manner. Interestingly, miRNAs also augmented expression of surface markers associated with NK cell maturation, such as CD56 and CD94, suggesting that miRNAs may play a role in promoting NK cell maturation. Mechanistically, we found that stimulation with miRNAs led to downstream activation of NF-kB. This effect was blunted upon blockade of TLR (e.g. TLR1) signaling, and was attenuated in lymphoma patients.Collectively, we provide the first evidence that extrinsic miRNAs, as a class of regulatory molecules, directly activate and may also promote the maturation of NK cells. These effects on NK cell activation and maturation are mediated, at least in part, by the TLR signaling pathway. This phenomenon may be important for normal host defense against infection and/or malignant transformation. Our studies indentify a new function of miRNAs with physiological relevance, and their potential for applications in preventing or treating cancer and infections either alone or as an adjuvant.Jaglowski: Pharmacyclics: Research Funding.

Abstract 2160

The mechanism underlying the robust expansion of natural killer (NK) cells during exogenous administration of FL is unknown. Endogenous IL-15 had no impact on the in vivo expansion of NK cell precursors during FL administration but was required for the FL-mediated expansion of mature NK cells in the spleen and blood. Studies performed using in vivo BM chimeras showed that cells derived from hematopoietic precursors (HPC), not stromal cells, provided the endogenous IL-15 required for mature NK cell expansion by FL administration. Exogenous administration of FL significantly increased both CD11b(+)CD11c(-) and CD11b(+)CD11c(+) populations but not their relatively abundant expression of IL-15 or IL-15 receptor alpha on a per cell basis. This increase preceded and correlated with NK cell expansion, the latter of which largely resulted from enhanced survival and proliferation of an existing pool of mature NK cells rather than increased de novo production of NK cells from bone marrow precursors. Finally, in vivo elimination of CD11c+ cells during the course of FL treatment significantly decreased NK cell expansion. In summary, FL administration increases NK cells in vivo by expanding antigen presenting cells which in turn provide the requisite IL-15 to enhance survival and proliferation of mature NK cells.

The mechanism underlying the robust expansion of natural killer (NK) cells during exogenous administration of FL is unknown. Endogenous IL-15 had no impact on the in vivoexpansion of NK cell precursors during FL administration but was required for the FL-mediated expansion of mature NK cells in the spleen and blood. Studies performed using in vivoBM chimeras showed that cells derived from hematopoietic precursors (HPC), not stromal cells, provided the endogenous IL-15 required for mature NK cell expansion by FL administration. Exogenous administration of FL significantly increased both CD11b(+)CD11c(-) and CD11b(+)CD11c(+) populations but not their relatively abundant expression of IL-15 or IL-15 receptor alpha on a per cell basis. This increase preceded and correlated with NK cell expansion, the latter of which largely resulted from enhanced survival and proliferation of an existing pool of mature NK cells rather than increased de novoproduction of NK cells from bone marrow precursors. Finally, in vivoelimination of CD11c+ cells during the course of FL treatment significantly decreased NK cell expansion. In summary, FL administration increases NK cells in vivoby expanding antigen presenting cells which in turn provide the requisite IL-15 to enhance survival and proliferation of mature NK cells.

Transforming growth factor-ß-stimulated clone-22 (TSC-22) is a gene that has been shown to be silenced in brain and prostate cancer, but its function and the mechanism responsible for this silencing are unknown. We used our model of spontaneous T-natural killer (NK) acute lymphoblastic leukemia (ALL) and discovered that the TSC-22 promoter was methylated resulting in absent expression in seven of eight cases of primary NK-T ALL, but not in cells from normal mice or mice with polyclonal expansion of T and NK cells. We found that TSC-22 was undetectable or minimally expressed in mouse lymphoma cell lines YAC-1 and EL-4 and human leukemia cell lines Jurkat and RPMI 8866, but treatment with the demethylation agent 5-aza-2'-deoxycytidine restored or increased TSC-22 expression. We mapped the TSC-22 promoter and discovered a CPG island in the proximal region and determined that its methylation was responsible for the decreased gene expression. Over-expression of TSC-22 slowed in vitro cell growth and resulted in a dramatic decrease of tumor size in vivo. Finally, TSC-22 expression was found to be absent or substantially reduced in human chronic lymphocytic leukemia and acute myeloid leukemia compared to normal human tissue. Collectively, our data indicate that TSC-22 is silenced via DNA methylation within its proximal promoter, and this silencing appears to contribute to its function as a putative tumor suppressor gene in leukemia. Silencing of TSC-22 can be reversed by 5-aza-2'-deoxycytidine, recently approved by the FDA for the treatment of myelodysplastic syndrome.

Monokines (i.e. IL-12, IL-18 and IL-15) induce natural killer (NK) cells to produce interferon-? (IFN-?), which is critical for monocyte clearance of infectious pathogens and tumor surveillance. To identify new regulators of IFN-? production we performed oligonucleotide array analysis of unstimulated and IL-12- and IL-18-stimulated NK92 cells. Among the subset of mRNAs differentially regulated in monokine-stimulated cells, we found SET, a potent inhibitor of the protein phosphatase type 2A (PP2A). SET mRNA and/or protein levels were upregulated in IL-12/IL-18- and IL-12/IL-15-stimulated primary human NK cells. Interestingly, the SET protein is also selectively increased in the resting CD56bright NK subset, which is a potent producer of IFN-? relative to the CD56dim NK subset. To determine whether SET positively regulates IFN-? production by inhibiting PP2A activity, we employed RNAi and interfered with SET expression in NK92 cells. SET downregulation inhibited IFN-? secretion by IL-12/IL-18, IL-12/IL-15- or IL-15/IL-18-stimulated NK92 cells. By contrast, ectopic SET expression increased IFN-? production in monokine-stimulated NK92 and primary human NK cells. Because downregulation of SET augmented PP2A activity in NK92 cells, we sought to investigate whether pharmacologic activation of PP2A inhibits the ability of NK cells to produce IFN-?. Indeed, suppression of IFN-? expression and secretion was also observed upon treatment of NK92 and primary NK cells with 1,9-dideoxy-forskolin, a known inducer of PP2A activity. Accordingly, NK cells from mice treated with 1.9-dideoxy-forskolin produced less IFN-? in response to in vivo monokine stimulation than did NK cells from vehicle-treated mice. Mechanistically, activation of PP2A by SET knock-down or 1,9-dideoxy-forskolin treatment leads to inhibition of ERK1/2, p65RelA and STAT5 activity in monokine-stimulated NK cells. Because these signaling molecules are important for IFN-? production by monokine-stimulated NK cells, our results strongly suggest that monokine induction of SET expression in NK cells is essential for limiting PP2A activity that, otherwise, would negatively impact the ability of NK cells to produce and release optimal levels of IFN-?.

Transforming growth factor-β-stimulated clone-22 (TSC-22) is a gene that has been shown to be silenced in brain and prostate cancer, but its function and the mechanism responsible for this silencing are unknown. We used our model of spontaneous T-natural killer (NK) acute lymphoblastic leukemia (ALL) and discovered that the TSC-22promoter was methylated resulting in absent expression in seven of eight cases of primary NK-T ALL, but not in cells from normal mice or mice with polyclonal expansion of T and NK cells. We found that TSC-22was undetectable or minimally expressed in mouse lymphoma cell lines YAC-1 and EL-4 and human leukemia cell lines Jurkat and RPMI 8866, but treatment with the demethylation agent 5-aza-2′-deoxycytidine restored or increased TSC-22expression. We mapped the TSC-22promoter and discovered a CPG island in the proximal region and determined that its methylation was responsible for the decreased gene expression. Over-expression of TSC-22slowed in vitrocell growth and resulted in a dramatic decrease of tumor size in vivo. Finally, TSC-22expression was found to be absent or substantially reduced in human chronic lymphocytic leukemia and acute myeloid leukemia compared to normal human tissue. Collectively, our data indicate that TSC-22is silenced via DNA methylation within its proximal promoter, and this silencing appears to contribute to its function as a putative tumor suppressor gene in leukemia. Silencing of TSC-22can be reversed by 5-aza-2′-deoxycytidine, recently approved by the FDA for the treatment of myelodysplastic syndrome.

Monokines (i.e. IL-12, IL-18 and IL-15) induce natural killer (NK) cells to produce interferon-γ (IFN-γ), which is critical for monocyte clearance of infectious pathogens and tumor surveillance. To identify new regulators of IFN-γ production we performed oligonucleotide array analysis of unstimulated and IL-12- and IL-18-stimulated NK92 cells. Among the subset of mRNAs differentially regulated in monokine-stimulated cells, we found SET, a potent inhibitor of the protein phosphatase type 2A (PP2A). SET mRNA and/or protein levels were upregulated in IL-12/IL-18- and IL-12/IL-15-stimulated primary human NK cells. Interestingly, the SET protein is also selectively increased in the resting CD56brightNK subset, which is a potent producer of IFN-γ relative to the CD56dimNK subset. To determine whether SET positively regulates IFN-γ production by inhibiting PP2A activity, we employed RNAi and interfered with SET expression in NK92 cells. SET downregulation inhibited IFN-γ secretion by IL-12/IL-18, IL-12/IL-15- or IL-15/IL-18-stimulated NK92 cells. By contrast, ectopic SET expression increased IFN-γ production in monokine-stimulated NK92 and primary human NK cells. Because downregulation of SET augmented PP2A activity in NK92 cells, we sought to investigate whether pharmacologic activation of PP2A inhibits the ability of NK cells to produce IFN-γ. Indeed, suppression of IFN-γ expression and secretion was also observed upon treatment of NK92 and primary NK cells with 1,9-dideoxy-forskolin, a known inducer of PP2A activity. Accordingly, NK cells from mice treated with 1.9-dideoxy-forskolin produced less IFN-γ in response to in vivomonokine stimulation than did NK cells from vehicle-treated mice. Mechanistically, activation of PP2A by SET knock-down or 1,9-dideoxy-forskolin treatment leads to inhibition of ERK1/2, p65RelA and STAT5 activity in monokine-stimulated NK cells. Because these signaling molecules are important for IFN-γ production by monokine-stimulated NK cells, our results strongly suggest that monokine induction of SET expression in NK cells is essential for limiting PP2A activity that, otherwise, would negatively impact the ability of NK cells to produce and release optimal levels of IFN-γ.

A tight control of kinase and phosphatase activity is fundamental for normal cell growth, survival and differentiation. The deregulated kinase activity of the BCR/ABL oncoprotein is responsible for the emergence and maintenance of chronic myelogenous leukemia (CML). By contrast, PP2A, a serine-threonine phosphatase involved in the regulation of many cellular functions, was found genetically inactivated in many types of cancer. We show here that, in BCR/ABL-transformed cells and CD34+ CML blast crisis progenitors, the phosphatase activity of the tumor suppressor PP2A is inhibited by the physiological PP2A-inhibitor SET whose expression is enhanced by BCR/ABL and increased in blast crisis CML. In imatinib-sensitive and -resistant (T315I included) BCR/ABL+ cell lines and in CD34+ CML blast crisis cells, molecular and/or pharmacological activation of PP2A leads to dephosphorylation of important regulators of proliferation and survival of CML progenitors, suppresses BCR/ABL kinase activity and promotes BCR/ABL proteasome degradation via a mechanism that requires the SHP-1 tyrosine phosphatase activity. Furthermore, PP2A activation achieved by shRNA-mediated SET knock-down or PP2Ac overexpression or treatment with the PP2A activator forskolin results in growth suppression, enhanced apoptosis, restored differentiation, impaired clonogenic potential and decreased in vivo leukemogenesis of wild type and T315I BCR/ABL-transformed myeloid cells. Thus, functional inactivation of PP2A phosphatase activity is essential for BCR/ABL leukemogenesis and, perhaps, required for transition of CML into blast crisis.

A tight control of kinase and phosphatase activity is fundamental for normal cell growth, survival and differentiation. The deregulated kinase activity of the BCR/ABL oncoprotein is responsible for the emergence and maintenance of chronic myelogenous leukemia (CML). By contrast, PP2A, a serine-threonine phosphatase involved in the regulation of many cellular functions, was found genetically inactivated in many types of cancer. We show here that, in BCR/ABL-transformed cells and CD34+CML blast crisis progenitors, the phosphatase activity of the tumor suppressor PP2A is inhibited by the physiological PP2A-inhibitor SET whose expression is enhanced by BCR/ABL and increased in blast crisis CML. In imatinib-sensitive and -resistant (T315I included) BCR/ABL+cell lines and in CD34+CML blast crisis cells, molecular and/or pharmacological activation of PP2A leads to dephosphorylation of important regulators of proliferation and survival of CML progenitors, suppresses BCR/ABL kinase activity and promotes BCR/ABL proteasome degradation via a mechanism that requires the SHP-1 tyrosine phosphatase activity. Furthermore, PP2A activation achieved by shRNA-mediated SET knock-down or PP2Ac overexpression or treatment with the PP2A activator forskolin results in growth suppression, enhanced apoptosis, restored differentiation, impaired clonogenic potential and decreased in vivoleukemogenesis of wild type and T315I BCR/ABL-transformed myeloid cells. Thus, functional inactivation of PP2A phosphatase activity is essential for BCR/ABL leukemogenesis and, perhaps, required for transition of CML into blast crisis.

Molecular characterization of human natural killer (NK) cells will require targeted gene delivery to inhibit and activate specific signaling pathways, yet to our knowledge, an effective means to deliver such products for long-term gene expression without disrupting normal cellular processes has not been described. In this study we have developed a retroviral strategy to effectively express gene products in the NK cell, whereby its effector functions of cytotoxicity and cytokine production remain intact. Using an EBV/retroviral hybrid vector PINCO, we demonstrate infection of human peripheral blood NK cells with simultaneous expression of a marker for infection - the enhanced green fluorescent protein (EGFP) - along with various genes of interest. This technique results in successful infection of the CD56dim NK population that predominates among human peripheral blood NK and is the effector of antibody-dependent cellular cytotoxicity (ADCC) and natural killing. In addition, we demonstrate infection of the CD56bright NK subset as well as the NK-92 and NK-L cell lines. Finally, we modify PINCO to express a cytoplasmically truncated murine CD8 molecule in place of GFP. The resulting vector enables us to transduce NK cells with multiple genes of interest simultaneously and provides an alternative purification method to FACS by using magnetic beads. In summary, we have devised an efficient and highly reproducible methodology for the targeted delivery of gene products to human NK cells that should now provide opportunities to dissect the molecular processes critical to normal NK cell physiology and to genetically manipulate NK cell populations prior to their administration in cancer therapy.

In adult humans, T cells differentiate in the thymus and B cells develop in the bone marrow, but the site(s) of natural killer (NK) cell differentiation are unclear. Here we describe, for the first time, a unique CD34(+) population found in human lymph nodes (LN) that differentiates into NK cells. CD56brightNK cells represent <10% of NK cells in peripheral blood (PB) yet predominate in LN where they can compete for endogenous T cell-derived IL-2 during immune activation due to their unique expression of functional high affinity (HA) interleukin (IL)-2 receptors (IL-2R). We hypothesized that a subset of CD34(+) hematopoietic precursor cells (HPC) might also express functional HA IL-2R and potentially differentiate into CD56brightNK cells via activation with low dose IL-2. We first identified a novel human CD34dimCD45RA(+) HPC in PB with constitutive expression of the HA IL-2R. When cultured in picomolar concentrations of IL-2 that selectively saturate the HA IL-2R, these cells give rise to CD56brightNK cells, and this effect is blocked when IL-2 cannot bind to its HA receptor. This unique CD34(+) population expresses IL-2Rα, CD2, CD7, c-kit, L-selectin, and NKR-P1A, all of which are also expressed by CD56brightNK cells. Unique among total PB CD34(+) cells, this novel population displays high integrin α4β7expression. This attribute, in addition to its high L-selectin expression, suggested that these cells may traffic to LN where their progeny, CD56brightNK cells, represent the major NK subset. Indeed we found a distinct CD34dimCD45RA(+)α4β7brightpopulation that resides in the T cell rich regions of human LN, and when stimulated in vitrowith 10 pM IL-2, this cell gives rise to CD56brightNK cells. This novel population represents only ~6% of all PB CD34(+) HPC yet is the major if not exclusive CD34(+) subset in LN. While murine studies strongly support the notion that most if not all NK cells require IL-15 for their development, these new human data suggest a model for development of a minor human NK subset, the CD56brightNK cells, whereby CD34dimCD45RA(+)α4β7brightHPC constitutively expressing the HA IL-2R traffic to peripheral LN where endogenous T cell-derived IL-2 can drive CD56brightNK cell differentiation in vivo.

In adult humans, T cells differentiate in the thymus and B cells develop in the bone marrow, but the site(s) of natural killer (NK) cell differentiation are unclear. Here we describe, for the first time, a unique CD34(+) population found in human lymph nodes (LN) that differentiates into NK cells. CD56bright NK cells represent <10% of NK cells in peripheral blood (PB) yet predominate in LN where they can compete for endogenous T cell-derived IL-2 during immune activation due to their unique expression of functional high affinity (HA) interleukin (IL)-2 receptors (IL-2R). We hypothesized that a subset of CD34(+) hematopoietic precursor cells (HPC) might also express functional HA IL-2R and potentially differentiate into CD56bright NK cells via activation with low dose IL-2. We first identified a novel human CD34dimCD45RA(+) HPC in PB with constitutive expression of the HA IL-2R. When cultured in picomolar concentrations of IL-2 that selectively saturate the HA IL-2R, these cells give rise to CD56bright NK cells, and this effect is blocked when IL-2 cannot bind to its HA receptor. This unique CD34(+) population expresses IL-2Rα, CD2, CD7, c-kit, L-selectin, and NKR-P1A, all of which are also expressed by CD56bright NK cells. Unique among total PB CD34(+) cells, this novel population displays high integrin α4β7 expression. This attribute, in addition to its high L-selectin expression, suggested that these cells may traffic to LN where their progeny, CD56bright NK cells, represent the major NK subset. Indeed we found a distinct CD34dimCD45RA(+)α4β7bright population that resides in the T cell rich regions of human LN, and when stimulated in vitro with 10 pM IL-2, this cell gives rise to CD56bright NK cells. This novel population represents only ~6% of all PB CD34(+) HPC yet is the major if not exclusive CD34(+) subset in LN. While murine studies strongly support the notion that most if not all NK cells require IL-15 for their development, these new human data suggest a model for development of a minor human NK subset, the CD56bright NK cells, whereby CD34dimCD45RA(+)α4β7bright HPC constitutively expressing the HA IL-2R traffic to peripheral LN where endogenous T cell-derived IL-2 can drive CD56bright NK cell differentiation in vivo.

Molecular characterization of human natural killer (NK) cells will require targeted gene delivery to inhibit and activate specific signaling pathways, yet to our knowledge, an effective means to deliver such products for long-term gene expression without disrupting normal cellular processes has not been described. In this study we have developed a retroviral strategy to effectively express gene products in the NK cell, whereby its effector functions of cytotoxicity and cytokine production remain intact. Using an EBV/retroviral hybrid vector PINCO, we demonstrate infection of human peripheral blood NK cells with simultaneous expression of a marker for infection - the enhanced green fluorescent protein (EGFP) - along with various genes of interest. This technique results in successful infection of the CD56dimNK population that predominates among human peripheral blood NK and is the effector of antibody-dependent cellular cytotoxicity (ADCC) and natural killing. In addition, we demonstrate infection of the CD56brightNK subset as well as the NK-92 and NK-L cell lines. Finally, we modify PINCO to express a cytoplasmically truncated murine CD8 molecule in place of GFP. The resulting vector enables us to transduce NK cells with multiple genes of interest simultaneously and provides an alternative purification method to FACS by using magnetic beads. In summary, we have devised an efficient and highly reproducible methodology for the targeted delivery of gene products to human NK cells that should now provide opportunities to dissect the molecular processes critical to normal NK cell physiology and to genetically manipulate NK cell populations prior to their administration in cancer therapy.