A novel LSD1 inhibitor NCD38 ameliorates MDS-related leukemia with complex karyotype by attenuating leukemia programs via activating super-enhancers

Lysine-specific demethylase 1 (LSD1) regulates gene expression by affecting histone modifications and is a promising target for acute myeloid leukemia (AML) with specific genetic abnormalities. Novel LSD1 inhibitors, NCD25 and NCD38, inhibited growth of MLL-AF9 leukemia as well as erythroleukemia, megakaryoblastic leukemia and myelodysplastic syndromes (MDSs) overt leukemia cells in the concentration range that normal hematopoiesis was spared. NCD25 and NCD38 invoked the myeloid development programs, hindered the MDS and AML oncogenic programs, and commonly upregulated 62 genes in several leukemia cells. NCD38 elevated H3K27ac level on enhancers of these LSD1 signature genes and newly activated ~500 super-enhancers. Upregulated genes with super-enhancer activation in erythroleukemia cells were enriched in leukocyte differentiation. Eleven genes including GFI1 and ERG, but not CEBPA, were identified as the LSD1 signature with super-enhancer activation. Super-enhancers of these genes were activated prior to induction of the transcripts and myeloid differentiation. Depletion of GFI1 attenuated myeloid differentiation by NCD38. Finally, a single administration of NCD38 causes the in vivo eradication of primary MDS-related leukemia cells with a complex karyotype. Together, NCD38 derepresses super-enhancers of hematopoietic regulators that are silenced abnormally by LSD1, attenuates leukemogenic programs and consequently exerts anti-leukemic effect against MDS-related leukemia with adverse outcome.

Principal component analysis uncovers cytomegalovirus-associated NK cell activation in Ph+ leukemia patients treated with dasatinib

Dasatinib treatment markedly increases the number of large granular lymphocytes (LGLs) in a proportion of Ph⁺ leukemia patients, which associates with a better prognosis. The lymphocytosis is predominantly observed in cytomegalovirus (CMV)-seropositive patients, yet detectable CMV reactivation exists only in a small fraction of patients. Thus, etiology of the lymphocytosis still remains unclear. Here, we identified NK cells as the dominant LGLs expanding in dasatinib-treated patients, and applied principal component analysis (PCA) to an extensive panel of NK cell markers to explore underlying factors in NK cell activation. PCA displayed phenotypic divergence of NK cells that reflects CMV-associated differentiation and genetic differences, and the divergence was markedly augmented in CMV-seropositive dasatinib-treated patients. Notably, the CMV-associated highly differentiated status of NK cells was already observed at leukemia diagnosis, and was further enhanced after starting dasatinib in virtually all CMV-seropositive patients. Thus, the extensive characterization of NK cells by PCA strongly suggests that CMV is an essential factor in the NK cell activation, which progresses stepwise during leukemia and subsequent dasatinib treatment most likely by subclinical CMV reactivation. This study provides a rationale for the exploitation of CMV-associated NK cell activation for treatment of leukemias.

Human herpesvirus-6 encephalitis during hematopoietic stem cell transplantation leads to poor prognosis

INTRODUCTION

Indications for the application of hematopoietic stem cell transplantation (HSCT) from alternative donors have remarkably broadened in scope; however, the incidence of infections that lead to failure of HSCT, such as human herpesvirus-6 (HHV-6) encephalitis, has also increased.

METHODS

We analyzed risk factors for symptomatic HHV-6 reactivation and the development of HHV-6 encephalitis in 140 consecutive adult patients who received allogeneic HSCT at our institution. Stem cell sources for the recipients were as follows: related-donor bone marrow in 40, related-donor peripheral blood in 5, unrelated bone marrow in 67, and unrelated cord blood in 28.

RESULTS

Symptomatic HHV-6 reactivation occurred in 22 patients (16%), and 11 patients manifested encephalitis. Multivariate Cox proportional hazards regression analysis identified cord blood cell transplantation (CBT) as an independent predictor of HHV-6 reactivation (P = 0.008). Hyponatremia or hypernatremia at the time of HHV-6 reactivation was detected before the development of HHV-6 encephalitis in 2 or 4 patients, respectively. Two patients died of HHV-6 encephalitis and 6 patients died of relapse of underlying diseases. Survival analysis identified higher risk of the disease (P = 0.021) and HHV-6 encephalitis (P = 0.003) as independent risk factors for reduced overall survival.

CONCLUSION

In cases involving CBT or unrelated-donor transplantation, patients should be carefully monitored for the symptomatic reactivation of HHV-6.

Control of an outbreak of carbapenem-resistant Pseudomonas aeruginosa in a haemato-oncology unit

An outbreak of a multidrug-resistant Pseudomonas aeruginosa producing metallo-β-lactamase (MBLPA) in a haemato-oncology unit was controlled using multidisciplinary interventions. The present study assesses the effects of these interventions by active surveillance of the incidence of MBLPA infection at the 1,240-bed tertiary care Kyoto University Hospital in Kyoto, Japan. Infection control strategies in 2004 included strengthening contact precautions, analysis of risk factors for MBLPA infection and cessation of urine collection. However, new MBLPA infections were identified in 2006, which prompted enhanced environmental cleaning, routine active surveillance, and restricting carbapenem usage. Between 2004 and 2010, 17 patients in the unit became infected with indistinguishable MBLPA strains. The final five infected patients were found by routine active surveillance, but horizontal transmission was undetectable. The MBLPA outbreak in the haemato-oncology unit was finally contained in 2008.

Replacement of the V3 region of gp120 with SDF-1 preserves the infectivity of T-cell line-tropic human immunodeficiency virus type 1

Interaction between the human immunodeficiency virus type 1 (HIV-1) envelope and the relevant chemokine receptors is crucial for subsequent membrane fusion and viral entry. Although the V3 region of gp120 is known to determine the cell tropism as well as the coreceptor usage, the significance of the binding of the V3 region to the chemokine receptor has not been fully understood. To address this issue, we adopted the pseudotyped virus infection assay in which the V3 region of the T-cell line-tropic (T-tropic) NL4-3 envelope was replaced with a portion of stromal cell-derived factor 1 (SDF-1), the ligand of CXCR4. The V3 region of the NL4-3 envelope expression vector was replaced with three different stretches of SDF-1 cDNA. Expression of each chimeric envelope protein was confirmed by immunoprecipitation and Western blotting. Luciferase reporter viruses were prepared by cotransfection of the pNL4-3.Luc.E(-)R(-) vector and each chimeric envelope expression vector, and the infection assay was then carried out. We showed that pseudotyped viruses with one of the chimeric envelopes, NL4-3/SDF1-51, could infect U87.CD4.CXCR4 but not U87.CD4 or U87.CXCR4 cells and that this infection was inhibited by the ligand of CXCR4, SDF-1beta, by anti-human SDF-1 antibody, or by an anti-CD4 antibody, Leu3a, in a dose-dependent manner. Furthermore, chimeric NL4-3/SDF1-51 gp120 significantly inhibited binding of labeled SDF-1 to CXCR4. It was suggested that replacement of the V3 region of the NL4-3 envelope with SDF-1 preserved the CD4-dependent infectivity of T-tropic HIV-1. These results indicate that binding between the V3 region and the relevant coreceptor is important for viral entry, whether its amino acid sequence is indigenous to the virus or not.

Molecular cloning of a novel human protein kinase, kpm, that is homologous to warts/lats, a Drosophila tumor suppressor

A novel human protein kinase, designated kpm, was identified and molecularly cloned. The isolated cDNA clone had an open reading frame consisting of 1088 amino acid residues with a putative kinase domain located near the carboxy-terminus. Homology search revealed that kpm belongs to a subfamily of serine/threonine protein kinases including warts/lats, a Drosophila tumor suppressor. Among these, kpm is most homologous to, but distinct from, recently reported LATS1, a human homolog of Drosophila warts/lats. Northern blot analysis disclosed that kpm is expressed as a 6.0 kb transcript in most of the tissues examined and also as an additional shorter 4.0 kb transcript in testis. Western blotting using polyclonal rabbit anti-kpm antibody detected kpm protein as a band with an apparent Mr of 150 kD. Immune complex kinase assay of HA-tagged kpm showed that kpm had kinase activity and phosphorylated itself in vitro. Studies with synchronized HeLa cells indicated that kpm protein was expressed relatively constantly throughout the cell cycle and underwent significant phosphorylation at mitotic phase. These results suggest that kpm plays a role in cell cycle progression during mitosis and its deletion or dysfunction might be involved in certain types of human cancers.

Both amino- and carboxyl-terminal domains of TRAF3 negatively regulate NF-kappaB activation induced by OX40 signaling

OX40 is a member of the tumor necrosis factor receptor (TNF-R) superfamily. We observed that overexpression of OX40 activated NF-kappaB, which was inhibited by dominant negative forms of TRAF2, NF-kappaB-inducing kinase (NIK), and IkappaB kinase (IKK) alpha. This indicates that OX40 signaling leads to NF-kappaB activation through the same cascade as TNF-R2. We then investigated the negative regulatory function of TRAF3 on OX40-induced NF-kappaB activation. TRAF3 blocked OX40-, TRAF2-induced NF-kappaB activation, but not NIK- and IKKalpha-induced NF-kappaB activation, indicating that TRAF3 blocks the pathway between TRAF2 and NIK. C-terminal deletion mutants as well as the N-terminal deletion mutant of TRAF3 inhibited NF-kappaB activation induced by OX40 or TRAF2. Since TRAF3 bound to OX40 through the C-terminal TRAF domain, the C-terminal domain is likely to work as a dominant negative mutant to compete the recruitment of TRAF2 to the receptor, which transmits the signal from OX40 to the downstream, NIK kinase. On the other hand, the N-terminal domain of TRAF3 seems to affect the downstream of TRAF2 binding. Thus, it is suggested that TRAF3 actively inhibits NF-kappaB activation induced by OX40.

Parathyroid hormone-related protein-induced hypercalcemia in SCID mice engrafted with adult T-cell leukemia cells

Parathyroid hormone-related protein (PTHrP) is considered to be one of the main causes of hypercalcemia associated with adult T-cell leukemia (ATL). To clarify the role of PTHrP and bone remodeling in the development of hypercalcemia in ATL, we examined the SCID mouse model of ATL that has previously been shown to mimic the disease in humans. Using this model, we found clear elevations in serum levels of calcium and C-terminal PTHrP (C-PTHrP). PTHrP mRNA was highly expressed in ATL cells proliferating in vivo. After the development of hypercalcemia, ATL mice were killed and bone histomorphometric analysis was performed. Bone volume was clearly decreased in the ATL mice. In comparison to control SCID mice, bone formation indices were very low in the ATL mice. Surprisingly, no significant difference was detected between the ATL mice and the control SCID mice in eroded surface/bone surface (ES/BS), a parameter of bone resorption. To our knowledge, the model presented here is the first animal model of ATL with humoral hypercalcemia. This is in contrast to previously reported, well-characterized animal models of human solid tumors associated with humoral hypercalcemia of malignancy (HHM). Furthermore, this model not only provides us with the opportunity to study the mechanisms underlying development of elevated calcium levels in ATL, but also allows us to test new therapeutic agents designed to treat hypercalcemia.

Activation of OX40 signal transduction pathways leads to tumor necrosis factor receptor-associated factor (TRAF) 2- and TRAF5-mediated NF-kappaB activation

We investigated the intracellular signaling of OX40, a member of the tumor necrosis factor receptor family. Activation of NF-kappaB in OX40-transfected HSB-2 cells was detected by electrophoretic mobility shift assay within 30 min after the binding of the ligand gp34. In vitro binding experiments showed that tumor necrosis factor receptor-associated factor (TRAF) 1, TRAF2, TRAF3, and TRAF5 but not TRAF4 associated with glutathione S-transferase-OX40 fusion protein. The cotransfection experiments using human embryo kidney cell derived HEK 293T cells showed that TRAF2, TRAF3, and TRAF5 associated with OX40 in vivo. Studies with OX40 deletion mutants demonstrated that the cytoplasmic portion consisting of amino acid sequence 256-263 (GGSFRTPI) was required for the association with TRAFs and NF-kappaB activation. The introduction of the dominant negative mutants of TRAF2 and TRAF5 into HSB-2-OX40 cells suppressed NF-kappaB activation in a dose-dependent manner. In addition, the introduction of TRAF3 together with the dominant negative mutants of TRAF2 or TRAF5 further reduced NF-kappaB activation. These results indicate that the NF-kappaB activation resulting from OX40 stimulation is mediated by both TRAF2 and TRAF5, and is likely to be negatively modulated by TRAF3.

Serial transplantation of adult T cell leukemia cells into severe combined immunodeficient mice

The precise mechanism of the neoplastic cell growth of adult T cell leukemia (ATL) still remains unclear. In the present study, we have succeeded in serial transplantation of ATL cells from a patient into severe combined immunodeficient (SCID) mice. In this model, we found that only a leukemic cell clone from an ATL patient could be successively transplanted into SCID mice, although it was difficult to maintain leukemic cell clones in vitro, suggesting that the microenvironment provided by SCID mice is suitable for leukemic cell growth. We could not detect human T cell leukemia virus type I (HTLV-I) mRNA or interleukin 2 (IL-2) mRNA in either the tumor cells growing in mice or the original leukemic cells. Thus, it appears that neither HTLV-I viral expression nor the IL-2 autocrine mechanism is directly involved in the neoplastic cell growth of fresh ATL cells as well as HTLV-I-infected cell lines, at least in SCID mice. In addition, we could passage frozen cells and obtain a large number of expanded leukemic cells in this model. Such a serial transplantation model, which can avoid the changes in the nature of leukemic cells that are frequently observed in in vitro culture, and which can propagate leukemic cell clones, would be very suitable not only to study the mechanism of neoplastic cell growth, but also to test potential therapeutic agents for ATL.

Radioimmunodetection of human leukemia with anti-interleukin-2 receptor antibody in severe combined immunodeficiency mice

Anti-Tac monoclonal antibody recognizes human interleukin-2 receptor, which is overexpressed in leukemic cells of most adult T-cell leukemia (ATL) patients. To examine the potency of anti-Tac for targeting of ATL, biodistributions of intravenously administered 125I- and 111In-labeled anti-Tac were examined in severe combined immunodeficiency (SCID) mice inoculated with ATL cells. Significant amounts of radiolabeled anti-Tac were found in the spleen and thymus. The trafficking of ATL cells in SCID mice was detected using 111In-oxine-labeled ATL cells. These results were coincident with the histologically confirmed infiltration of ATL cells. The radiolabeled anti-Tac seemed potent for targeting of ATL.

Tumorigenicity of human T-cell leukemia virus type I-infected cell lines in severe combined immunodeficient mice and characterization of the cells proliferating in vivo

The mechanism involved in leukemogenesis and neoplastic cell growth of adult T-cell leukemia (ATL) still remains unclear. We examined the tumorigenicity of human T-cell leukemia virus type I (HTLV-I)-infected cell lines in an in vivo cell proliferation model using severe combined immunodeficient (SCID) mice. Eleven HTLV-I-infected cell lines were injected into SCID mice and we found that 4 of them were capable of proliferating in SCID mice. Three of four transplantable cell lines are derived from the leukemic cell clone and 6 of 6 HTLV-I-infected cell lines of nonleukemic cell origin could not engraft in SCID mice. Interestingly, it was shown that some HTLV-I-infected and interleukin-2 (IL-2)-dependent cell lines could successfully engraft in SCID mice. The expression of IL-2 mRNA was not detected in these cell lines growing either in vivo or in vitro. HTLV-I viral products were not detected in 3 of 4 transplantable cell lines proliferating in vivo. Peripheral blood T cells immortalized by introduction of tax gene of HTLV-I were found to have no tumorigenic potential in SCID mice. These data suggest that (1) HTLV-I-infected cell lines of nonleukemic cell origin do not have enough leukemogenic changes to acquire the tumorigenic potential in SCID mice; (2) the IL-2 autocrine mechanism is not directly involved in the tumor cell growth; (3) viral gene expression is not needed for the maintenance of neoplastic cell growth; and (4) the expression of tax gene is not sufficient for the neoplastic cell growth in vivo.

[Analysis of in vivo cell proliferation of ATL using SCID mice]

We made a model of in vivo cell proliferation of leukemic cells from adult T cell leukemia (ATL) patients using severe combined immunodeficient (SCID) mice. SCID mice injected with ATL cells from 6 of 8 ATL patients were found to have the tumor. DNA analysis revealed that the clone of the cells proliferating in mice was the same as that of the original leukemic cells. Histologic examination showed that the pattern of the infiltration of ATL cells in mice was similar to that of an ATL patient. Next, we examined the tumorigenicity of HTLV-I infected cell lines using SCID mice. Seven HTLV-I infected cell lines were injected into SCID mice and it was found that 4 of them were capable of proliferating in SCID mice. HTLV-I infected cell lines of non-leukemic cell origin could not engraft in SCID mice, indicating that these cells seemed not to have the enough genetic changes to acquire the tumorigenic potential. Analysis of gene expression suggested that neither IL-2 nor HTLV-I viral product was directly involved in the neoplastic cell growth of ATL. Furthermore, T cells immortalized by introduction of Tax could not engraft in SCID mice, indicating that the expression of tax gene seemed not to be sufficient for the neoplastic cell growth in vivo.

Detection of homing, proliferation, and infiltration sites of adult T cell leukemia cells in severe combined immunodeficiency mice using radiometric techniques

To clarify the mechanism of in vivo proliferation of adult T cell leukemia (ATL) cells, we examined the organ distribution of ATL-43T cell line cells derived from original leukemic cells in severe combined immunodeficiency (SCID) mice using radiometric techniques. First, we injected 111In-oxine-labeled ATL-43T cells into SCID and CB17 mice. On day 6, significant accumulation of radioactivity was found in the spleen and thymus of SCID mice (33.3 +/- 9.4 and 10.0 +/- 3.6% injected dose/g of tissue [%ID/g], respectively) in comparison with that in CB17 mice (19.1 +/- 2.5 and 3.7 +/- 0.9%ID/g, respectively). Next, we injected radiolabeled anti-Tac monoclonal antibody (MoAb) recognizing human interleukin-2 receptor (IL-2R) alpha chain or isotype-matched control MoAb RPC5 in SCID mice bearing ATL-43T cells 4 weeks after cell inoculation. The amounts of radioactivity found in the spleen and thymus of SCID mice injected with 125I-labeled anti-Tac MoAb (22.5 +/- 6.9 and 22.8 +/- 9.6 %ID/g, respectively) were significantly higher than those in the corresponding organs of SCID mice injected with 125I-labeled RPC5 MoAb (12.0 +/- 5.1 and 7.5 +/- 4.6 %ID/g, respectively). Similar results were obtained with 111In-labeled anti-Tac MoAb. These results were consistent with the histological findings of SCID mice bearing ATL-43T cells, indicating that ATL-43T cells infiltrated preferentially into the lymphoid organs, such as the spleen and thymus, and proliferated there. Thus, the radiometric techniques employed in this study were very useful to evaluate the proliferation sites of ATL-43T cells in SCID mice. Furthermore, this murine model could give us an opportunity to test the feasibility of therapeutic application of radiolabeled anti-Tac MoAb.

[Usefulness of the measurement of serum soluble IL-2 receptor alpha chain levels in clinical monitoring of non-Hodgkin lymphoma]

We measured the serum concentrations of the soluble form of interleukin-2 receptor alpha chain (sIL-2R alpha) in 25 patients with non-Hodgkin lymphoma (NHL) and 1 patient with virus-associated hemophagocytic syndrome (VHAS) by an enzyme-linked immunosorbent assay (ELISA) using two anti-IL-2R alpha monoclonal antibodies which recognize different epitopes. The sIL-2R alpha levels were markedly higher (n = 12, range = 469.2-11020 U/ml, mean +/- SD = 6153.0 +/- 1687.2 U/ml) in the sera of patients with NHL than in healthy individuals (n = 46, range = 290.1-849.3 U/ml, mean +/- SD = 459.8 +/- 126.9 U/ml, p < 0.01). The serum sIL-2R alpha levels in the NHL patients were closely associated with the stage of disease. The mean value of the serum sIL-2R alpha in each stage of NHL was as follows: stage I (469.2 U/ml, n = 1), stage II (4879.0 U/ml, n = 3), stage III (7364.8 U/ml, n = 8), stage IV (13796.2 U/ml, n = 10). However, there was no clear correlation between the elevation of serum sIL-2R alpha level and the pathological (LSG) classification of NHL. The serum sIL-2R alpha levels serially measured during the clinical course of 2 NHL patients were closely associated with the severity and the progression of the disease. Thus, the measurement of serum sIL-2R alpha levels is very useful not only for the diagnosis but also to monitor the clinical course of NHL.