Targeted cytotoxic effect of anti-JL1 immunotoxin against a human leukemic cell line and its clinical implications

We have previously reported the identification of a unique thymocyte-specific surface molecule, JL1, which was detected using the monoclonal antibody (mAb), anti-JL1. Interestingly, JL1 was shown to be expressed in most leukemias, irrespective of their immunophenotype, and subpopulations of normal bone marrow (BM) mononuclear cells (MNCs). Here we investigated the potential usefulness of the anti-JL1 mAb as a therapeutic tool for leukemia. We demonstrated that the proliferation of cultured human leukemia cells was dramatically inhibited in vitro by anti-JL1 mAb conjugated with the polypeptide toxin, gelonin, but not by gelonin alone. We then systematically investigated the reactivity of the anti-JL1 mAb against normal human tissues to evaluate possible side effects along with various hematopoietic and nonhematopoietic tumor cell lines. All of 33 types of normal tissues except thymus and subpopulation of BM MNCs were clearly devoid of JL1 expression. Among tumor cell lines, all the nonhematopoietic cell lines tested were negative for JL1 expression, while some hematopoietic cell lines contained JL1 antigen. Collectively, the results showed the cytotoxic effects of anti-JL1-based immunotoxin against JL1-positive leukemic cells, sparing most normal tissues other than thymocytes and some BM MNCs. Therefore, we strongly suggest that gelonin-conjugated anti-JL1 mAb immunotoxin could be developed as a potential immunotherapeutic agent in the treatment of various types of JL1-positive acute leukemias.     

Leukemia-specific siRNA delivery by immunonanoplexes consisting of anti-JL1 minibody conjugated to oligo-9 Arg-peptides

Targeted mRNA degradation by short interfering RNAs (siRNAs) offers a great potential to treat cancers. siRNA therapeutics for leukemias are, however, hindered by poor intracellular uptake, limited blood stability and nonspecific delivery. To solve these problems, we developed an anti-JL1 immunonanoplex (antibody-coupled nanocomplex) for siRNA delivery using anti-JL1 minibody (leukemia cell-specific minibody) conjugated to oligo-9-Arg peptide (9R) for effective siRNA delivery to leukemic cells. The anti-JL1 immunonanoplexes were able to deliver siRNA specifically to leukemic cells (CEM and Jurkat), but not to control cancer cells (H9). According to FACS and confocal microscopic analysis, siRNAs delivered by immunonanoplex particles were rapidly taken up by the JL1-positive cancer cells in 2 h. Furthermore, we showed that the anti-JL1 immunonanoplexes were effectively targeted to JL1-positive cells (CEM) inoculated in the mouse bone marrow. These results suggest that the anti-JL1 immunonanoplex is a powerful siRNA delivery system for human leukemia therapies.     

p53 restoration kills primitive leukemia cells in vivo and increases survival of leukemic mice

Loss of p53 function is a common feature of human cancers and it is required for differentiated tumor cell maintenance; however, it is not known whether sustained inactivation of the p53 pathway is needed for cancer stem cell persistence. Chronic myeloid leukemia (CML) is caused by a chromosome translocation that generates the BCRABL oncogene encoding a constitutively active protein tyrosine kinase. The disease originates in a hematopoietic stem cell and is maintained by leukemic stem cells (LSCs). Treatment with specific tyrosine kinase inhibitors does not eliminate LSCs because they do not depend on the oncogene for survival. We have combined a switchable p53 knock-in mouse model, p53^KI/KI, with the well-characterized Sca1-BCRABLp210 CML transgenic model, to show that transient restoration of p53 slows disease progression and significantly extends the survival of leukemic animals, being the mechanism responsible for this effect, apoptotic death of primitive leukemia cells. In agreement with these in vivo findings, in vitro assays show that restoring p53 reduces hematopoietic colony formation by cells of leukemic animals. These results suggest that reestablishing p53 function may be a therapeutic strategy for the eradication of leukemic stem cells and to prevent disease progression.     

An evidence for adhesion-mediated acquisition of acute myeloid leukemic stem cell-like immaturities

For long-term survival in vitro and in vivo of acute myeloid leukemia cells, their adhesion to bone marrow stromal cells is indispensable. However, it is still unknown if these events are uniquely induced by the leukemic stem cells. Here we show that TF-1 human leukemia cells, once they have formed a cobblestone area by adhering to mouse bone marrow-derived MS-5 cells, can acquire some leukemic stem cell like properties in association with a change in the CD44 isoform-expression pattern and with an increase in a set of related microRNAs. These findings strongly suggest that at least some leukemia cells can acquire leukemic stem cell like properties in an adhesion-mediated stochastic fashion.     

Inhibition of HER3 activation and tumor growth with a human antibody binding to a conserved epitope formed by domain III and IV

Human epidermal growth factor receptor 3 (HER3, also known as ErbB3) has emerged as relevant target for antibody-mediated tumor therapy. Here, we describe a novel human antibody, IgG 3–43, recognizing a unique epitope formed by domain III and parts of domain IV of the extracellular region of HER3, conserved between HER3 and mouse ErbB3. An affinity of 11 nM was determined for the monovalent interaction. In the IgG format, the antibody bound recombinant bivalent HER3 with subnanomolar affinity (K_D = 220 pM) and HER3-expressing tumor cells with EC₅₀ values in the low picomolar range (27 - 83 pM). The antibody competed with binding of heregulin to HER3-expressing cells, efficiently inhibited phosphorylation of HER3 as well as downstream signaling, and induced receptor internalization and degradation. Furthermore, IgG 3–43 inhibited heregulin-dependent proliferation of several HER3-positive cancer cell lines and heregulin-independent colony formation of HER2-overexpressing tumor cell lines. Importantly, inhibition of tumor growth and prolonged survival was demonstrated in a FaDu xenograft tumor model in SCID mice. These findings demonstrate that by binding to the membrane-proximal domains III and IV involved in ligand binding and receptor dimerization, IgG 3–43 efficiently inhibits activation of HER3, thereby blocking tumor cell growth both in vitro and in vivo.     

Targeting calcineurin activation as a therapeutic strategy for T-cell acute lymphoblastic leukemia

Calcineurin is a calcium-activated serine/threonine phosphatase critical to a number of developmental processes in the cardiovascular, nervous and immune systems. In the T-cell lineage, calcineurin activation is important for pre-T-cell receptor (TCR) signaling, TCR-mediated positive selection of thymocytes into mature T cells, and many aspects of the immune response. The critical role of calcineurin in the immune response is underscored by the fact that calcineurin inhibitors, such as cyclosporin A (CsA) and FK506, are powerful immunosuppressants in wide clinical use. We observed sustained calcineurin activation in human B- and T-cell lymphomas and in all mouse models of lymphoid malignancies analyzed. In intracellular NOTCH1 (ICN1)- and TEL-JAK2-induced T-cell lymphoblastic leukemia, two mouse models relevant to human malignancies, in vivo inhibition of calcineurin activity by CsA or FK506 induced apoptosis of leukemic cells and rapid tumor clearance, and substantially prolonged mouse survival. In contrast, ectopic expression of a constitutively activated mutant of calcineurin favored leukemia progression. Moreover, CsA treatment induced apoptosis in human lymphoma and leukemia cell lines. Thus, calcineurin activation is critical for the maintenance of the leukemic phenotype in vivo, identifying this pathway as a relevant therapeutic target in lymphoid malignancies.     

Requirement for CD44 in homing and engraftment of BCR-ABL-expressing leukemic stem cells

In individuals with chronic myeloid leukemia (CML) treated by autologous hematopoietic stem cell (HSC) transplantation, malignant progenitors in the graft contribute to leukemic relapse, but the mechanisms of homing and engraftment of leukemic CML stem cells are unknown. Here we show that CD44 expression is increased on mouse stem-progenitor cells expressing BCR-ABL and that CD44 contributes functional E-selectin ligands. In a mouse retroviral transplantation model of CML, BCR-ABL1-transduced progenitors from CD44-mutant donors are defective in homing to recipient marrow, resulting in decreased engraftment and impaired induction of CML-like myeloproliferative disease. By contrast, CD44-deficient stem cells transduced with empty retrovirus engraft as efficiently as do wild-type HSCs. CD44 is dispensable for induction of acute B-lymphoblastic leukemia by BCR-ABL, indicating that CD44 is specifically required on leukemic cells that initiate CML. The requirement for donor CD44 is bypassed by direct intrafemoral injection of BCR-ABL1-transduced CD44-deficient stem cells or by coexpression of human CD44. Antibody to CD44 attenuates induction of CML-like leukemia in recipients. These results show that BCR-ABL-expressing leukemic stem cells depend to a greater extent on CD44 for homing and engraftment than do normal HSCs, and argue that CD44 blockade may be beneficial in autologous transplantation in CML.     

Mass spectrometry-based identification of the tumor antigen UN1 as the transmembrane CD43 sialoglycoprotein

The UN1 monoclonal antibody recognized the UN1 antigen as a heavily sialylated and O-glycosylated protein with the apparent molecular weight of 100-120 kDa; this antigen was peculiarly expressed in fetal tissues and several cancer tissues, including leukemic T cells, breast, and colon carcinomas. However, the lack of primary structure information has limited further investigation on the role of the UN1 antigen in neoplastic transformation. In this study, we have identified the UN1 antigen as CD43, a transmembrane sialoglycoprotein involved in cell adhesion, differentiation, and apoptosis. Indeed, mass spectrometry detected two tryptic peptides of the membrane-purified UN1 antigen that matched the amino acidic sequence of the CD43 intracellular domain. Immunological cross-reactivity, migration pattern in mono- and bi-dimensional electrophoresis, and CD43 gene-dependent expression proved the CD43 identity of the UN1 antigen. Moreover, the monosaccharide GalNAc-O-linked to the CD43 peptide core was identified as an essential component of the UN1 epitope by glycosidase digestion of specific glycan branches. UN1-type CD43 glycoforms were detected in colon, sigmoid colon, and breast carcinomas, whereas undetected in normal tissues from the same patients, confirming the cancer-association of the UN1 epitope. Our results highlight UN1 monoclonal antibody as a suitable tool for cancer immunophenotyping and analysis of CD43 glycosylation in tumorigenesis.     

Antibody-mediated phagocytosis contributes to the anti-tumor activity of the therapeutic antibody daratumumab in lymphoma and multiple myeloma

Daratumumab (DARA) is a human CD38-specific IgG1 antibody that is in clinical development for the treatment of multiple myeloma (MM). The potential for IgG1 antibodies to induce macrophage-mediated phagocytosis, in combination with the known presence of macrophages in the tumor microenvironment in MM and other hematological tumors, led us to investigate the contribution of antibody-dependent, macrophage-mediated phagocytosis to DARA's mechanism of action. Live cell imaging revealed that DARA efficiently induced macrophage-mediated phagocytosis, in which individual macrophages rapidly and sequentially engulfed multiple tumor cells. DARA-dependent phagocytosis by mouse and human macrophages was also observed in an in vitro flow cytometry assay, using a range of MM and Burkitt's lymphoma cell lines. Phagocytosis contributed to DARA's anti-tumor activity in vivo, in both a subcutaneous and an intravenous leukemic xenograft mouse model. Finally, DARA was shown to induce macrophage-mediated phagocytosis of MM cells isolated from 11 of 12 MM patients that showed variable levels of CD38 expression. In summary, we demonstrate that phagocytosis is a fast, potent and clinically relevant mechanism of action that may contribute to the therapeutic activity of DARA in multiple myeloma and potentially other hematological tumors.     

Augmentation of antileukemia lytic activity by OKT3 monoclonal antibody: synergism of OKT3 and interleukin-2

We have shown that short-term incubation (45 min) of peripheral blood lymphocytes of normal donors with OKT3 monoclonal antibody (MoAb), directed against T-cell-associated antigen CD3, resulted in an acquisition of lytic activity against fresh leukemic cells. Induction of such antileukemia activity was specific for OKT3, since Leu-1 MoAb (directed against another T cell surface molecule, CD5) did not induce a lytic effect. The OKT3-generated antileukemia effect was displayed against various types of leukemia including chronic myelogenous leukemia and acute myelogenous leukemia of various histological subtypes (M1, M2, M5). We furthermore demonstrated that OKT3 MoAb substantially enhanced leukemia killing by interleukin-2 (IL-2)-activated killer cells obtained from peripheral blood of patients with leukemia. Of most importance was the observation that the combined treatment of effector cells with IL-2 and OKT3 MoAb resulted in the highest levels of lysis of both autologous and allogeneic fresh leukemic cells that have been observed in leukemic patients to date. Of importance was to note that OKT3 treatment was effective in induction of cytotoxic activity also in patients whose effector cells were unresponsive to stimulation with IL-2 alone. All of these observations suggest that IL-2-activated and OKT3-MoAb-treated effector cells may represent the most aggressive population of antileukemia-directed killer cells and may play a significant role in the treatment of human leukemia.     

CD1d-antibody fusion proteins target iNKT cells to the tumor and trigger long-term therapeutic responses

Despite the well-established antitumor activity of CD1d-restricted invariant natural killer T lymphocytes (iNKT), their use for cancer therapy has remained challenging. This appears to be due to their strong but short-lived activation followed by long-term anergy after a single administration of the CD1d agonist ligand alpha-galactosylceramide (αGC). As a promising alternative, we obtained sustained mouse iNKT cell responses associated with prolonged antitumor effects through repeated administrations of tumor-targeted recombinant sCD1d-antitumor scFv fusion proteins loaded with αGC. Here, we demonstrate that CD1d fusion proteins bound to tumor cells via the antibody fragment specific for a tumor-associated antigen, efficiently activate human iNKT cell lines leading to potent tumor cell lysis. The importance of CD1d tumor targeting was confirmed in tumor-bearing mice in which only the specific tumor-targeted CD1d fusion protein resulted in tumor inhibition of well-established aggressive tumor grafts. The therapeutic efficacy correlated with the repeated activation of iNKT and natural killer cells marked by their release of TH1 cytokines, despite the up-regulation of the co-inhibitory receptor PD-1. Our results demonstrate the superiority of providing the superagonist αGC loaded on recombinant CD1d proteins and support the use of αGC/sCD1d-antitumor fusion proteins to secure a sustained human and mouse iNKT cell activation, while targeting their cytotoxic activity and cytokine release to the tumor site.     

Optical imaging of disseminated leukemia models in mice with near-infrared probe conjugated to a monoclonal antibody

Background

The assessment of anticancer agents to treat leukemia needs to have animal models closer to the human pathology such as implantation in immunodeficient mice of leukemic cells from patient samples. A sensitive and early detection of tumor cells in these orthotopic models is a prerequisite for monitoring engraftment of leukemic cells and their dissemination in mice. Therefore, we developed a fluorescent antibody based strategy to detect leukemic foci in mice bearing patient-derived leukemic cells using fluorescence reflectance imaging (FRI) to determine when to start treatments with novel antitumor agents.

Methods

Two mAbs against the CD44 human myeloid marker or the CD45 human leukocyte marker were labeled with Alexa Fluor 750 and administered to leukemia-bearing mice after having verified the immunoreactivity in vitro. Bioluminescent leukemic cells (HL60-Luc) were used to compare the colocalization of the fluorescent mAb with these cells. The impact of the labeled antibodies on disease progression was further determined. Finally, the fluorescent hCD45 mAb was tested in mice engrafted with human leukemic cells.

Results

The probe labeling did not modify the immunoreactivity of the mAbs. There was a satisfactory correlation between bioluminescence imaging (BLI) and FRI and low doses of mAb were sufficient to detect leukemic foci. However, anti-hCD44 mAb had a strong impact on the tumor proliferation contrary to anti-hCD45 mAb. The use of anti-hCD45 mAb allowed the detection of leukemic patient cells engrafted onto NOD/SCID mice.

Conclusions

A mAb labeled with a near infrared fluorochrome is useful to detect leukemic foci in disseminated models provided that its potential impact on tumor proliferation has been thoroughly documented.     

Targeting of CD44 eradicates human acute myeloid leukemic stem cells

The long-term survival of patients with acute myeloid leukemia (AML) is dismally poor. A permanent cure of AML requires elimination of leukemic stem cells (LSCs), the only cell type capable of initiating and maintaining the leukemic clonal hierarchy. We report a therapeutic approach using an activating monoclonal antibody directed to the adhesion molecule CD44. In vivo administration of this antibody to nonobese diabetic-severe combined immune-deficient mice transplanted with human AML markedly reduced leukemic repopulation. Absence of leukemia in serially transplanted mice demonstrated that AML LSCs are directly targeted. Mechanisms underlying this eradication included interference with transport to stem cell-supportive microenvironmental niches and alteration of AML-LSC fate, identifying CD44 as a key regulator of AML LSCs. The finding that AML LSCs require interaction with a niche to maintain their stem cell properties provides a therapeutic strategy to eliminate quiescent AML LSCs and may be applicable to other types of cancer stem cells.     

A Critical Role of CDKN3 in Bcr-Abl-Mediated Tumorigenesis

CDKN3 (cyclin-dependent kinase inhibitor 3), a dual specificity protein phosphatase, dephosphorylates cyclin-dependent kinases (CDKs) and thus functions as a key negative regulator of cell cycle progression. Deregulation or mutations of CDNK3 have been implicated in various cancers. However, the role of CDKN3 in Bcr-Abl-mediated chronic myelogenous leukemia (CML) remains unknown. Here we found that CDKN3 acts as a tumor suppressor in Bcr-Abl-mediated leukemogenesis. Overexpression of CDKN3 sensitized the K562 leukemic cells to imanitib-induced apoptosis and dramatically inhibited K562 xenografted tumor growth in nude mouse model. Ectopic expression of CDKN3 significantly reduced the efficiency of Bcr-Abl-mediated transformation of FDCP1 cells to growth factor independence. In contrast, depletion of CDKN3 expression conferred resistance to imatinib-induced apoptosis in the leukemic cells and accelerated the growth of xenograph leukemia in mice. In addition, we found that CDKN3 mutant (CDKN3-C140S) devoid of the phosphatase activity failed to affect the K562 leukemic cell survival and xenografted tumor growth, suggesting that the phosphatase of CDKN3 was required for its tumor suppressor function. Furthermore, we observed that overexpression of CDKN3 reduced the leukemic cell survival by dephosphorylating CDK2, thereby inhibiting CDK2-dependent XIAP expression. Moreover, overexpression of CDKN3 delayed G1/S transition in K562 leukemic cells. Our results highlight the importance of CDKN3 in Bcr-Abl-mediated leukemogenesis, and provide new insights into diagnostics and therapeutics of the leukemia.     

E4F1 dysfunction results in autophagic cell death in myeloid leukemic cells

The multifunctional E4F1 protein was originally identified as a cellular target of the E1A adenoviral oncoprotein. Although E4F1 is implicated in several key oncogenic pathways, its roles in tumorigenesis remain unclear. Using a genetically engineered mouse model of myeloid leukemia (histiocytic sarcomas, HS) based on the genetic inactivation of the tumor suppressor Ink4a/Arf locus, we have recently unraveled an unsuspected function of E4F1 in the survival of leukemic cells. In vivo, genetic ablation of E4F1 in established myeloid tumors results in tumor regression. E4F1 inactivation results in a cascade of alterations originating from dysfunctional mitochondria that induce increased reactive oxygen species (ROS) levels and ends in massive autophagic cell death in HS transformed, but not normal myeloid cells. E4F1 depletion also induces cell death in various human myeloid leukemic cell lines, including acute myeloid leukemic (AML) cell lines. Interestingly, the E4F1 protein is overexpressed in a large proportion of human AML samples. These data provide new insights into E4F1-associated survival functions implicated in tumorigenesis and could open the path for new therapeutic strategies.     

Glycosylation Status of CD43 Protein Is Associated with Resistance of Leukemia Cells to CTL-Mediated Cytolysis

To improve cancer immunotherapy, it is important to understand how tumor cells counteract immune-surveillance. In this study, we sought to identify cell-surface molecules associated with resistance of leukemia cells to cytotoxic T cell (CTL)-mediated cytolysis. To this end, we first established thousands of monoclonal antibodies (mAbs) that react with MLL/AF9 mouse leukemia cells. Only two of these mAbs, designated R54 and B2, bound preferentially to leukemia cells resistant to cytolysis by a tumor cell antigen–specific CTLs. The antigens recognized by these mAbs were identified by expression cloning as the same protein, CD43, although their binding patterns to subsets of hematopoietic cells differed significantly from each other and from a pre-existing pan-CD43 mAb, S11. The epitopes of R54 and B2, but not S11, were sialidase-sensitive and expressed at various levels on leukemia cells, suggesting that binding of R54 or B2 is associated with the glycosylation status of CD43. R54^high leukemia cells, which are likely to express sialic acid-rich CD43, were highly resistant to CTL-mediated cytolysis. In addition, loss of CD43 in leukemia cells or neuraminidase treatment of leukemia cells sensitized leukemia cells to CTL-mediated cell lysis. These results suggest that sialic acid-rich CD43, which harbors multiple sialic acid residues that impart a net negative surface charge, protects leukemia cells from CTL-mediated cell lysis. Furthermore, R54^high or B2^high leukemia cells preferentially survived in vivo in the presence of adaptive immunity. Taken together, these results suggest that the glycosylation status of CD43 on leukemia is associated with sensitivity to CTL-mediated cytolysis in vitro and in vivo. Thus, regulation of CD43 glycosylation is a potential strategy for enhancing CTL-mediated immunotherapy.     

Antibody-mediated phagocytosis contributes to the anti-tumor activity of the therapeutic antibody daratumumab in lymphoma and multiple myeloma

Daratumumab (DARA) is a human CD38-specific IgG1 antibody that is in clinical development for the treatment of multiple myeloma (MM). The potential for IgG1 antibodies to induce macrophage-mediated phagocytosis, in combination with the known presence of macrophages in the tumor microenvironment in MM and other hematological tumors, led us to investigate the contribution of antibody-dependent, macrophage-mediated phagocytosis to DARA''s mechanism of action. Live cell imaging revealed that DARA efficiently induced macrophage-mediated phagocytosis, in which individual macrophages rapidly and sequentially engulfed multiple tumor cells. DARA-dependent phagocytosis by mouse and human macrophages was also observed in an in vitro flow cytometry assay, using a range of MM and Burkitt''s lymphoma cell lines. Phagocytosis contributed to DARA''s anti-tumor activity in vivo, in both a subcutaneous and an intravenous leukemic xenograft mouse model. Finally, DARA was shown to induce macrophage-mediated phagocytosis of MM cells isolated from 11 of 12 MM patients that showed variable levels of CD38 expression. In summary, we demonstrate that phagocytosis is a fast, potent and clinically relevant mechanism of action that may contribute to the therapeutic activity of DARA in multiple myeloma and potentially other hematological tumors.     

Modulation of NF-kappa B activity and apoptosis in chronic lymphocytic leukemia B cells

Chronic lymphocytic leukemia (CLL) is an indolent malignancy of CD5+ B lymphocytes. CLL cells express CD40, a key regulator of B cell proliferation, differentiation, and survival. In nonmalignant B cells, CD40 ligation results in nuclear translocation and activation of NF-kappaB proteins. Based on observations that in some CLL cases, the tumor cells express both CD40 and its ligand, CD154 (CD40 ligand), we proposed a model for CLL pathogenesis due to CD40 ligation within the tumor. To evaluate this issue, we used freshly isolated CLL B cells to examine constitutive and inducible NF-kappaB activity by electrophoretic mobility shift assay. We consistently observed high levels of nuclear NF-kappaB-binding activity in unstimulated CLL B cells relative to that detected in nonmalignant human B cells. In each case examined, CD40 ligation further augmented NF-kappaB activity and prolonged CLL cell survival in vitro. The principle NF-kappaB proteins in stimulated CLL cells appear to be quite similar to those in nonmalignant human B cells and include p50, p65, and c-Rel. In a CD154-positive case, blocking CD154 engagement by mAb to CD154 resulted in inhibition of NF-kappaB activity in the CLL cells. The addition of anti-CD154 mAb resulted in accelerated CLL cell death to a similar degree as was observed in cells exposed to dexamethasone. These data indicate that CD40 engagement has a profound influence on NF-kappaB activity and survival in CLL B cells, and are consistent with a role for CD154-expressing T and B cells in CLL pathogenesis. The data support the development of novel therapies based on blocking the CD154-CD40 interaction in CLL.     

The growth of K562 human leukemia cells was inhibited by therapeutic neural stem cells in cellular and xenograft mouse models

To confirm the anti-tumor effect of engineered neural stem cells (NSCs) expressing cytosine deaminase (CD) and interferon-β (IFN-β) with prodrug 5-fluorocytosine (FC), K562 chronic myeloid leukemia (CML) cells were co-cultured with the neural stem cell lines HB1.F3.CD and HB1.F3.CD.IFN-β in 5-FC containing media. A significant decrease in the viability of K562 cells was observed by the treatment of the NSC lines, HB1.F3.CD and HB1.F3.CD.IFN-β, compared with the control. A modified trans-well assay showed that engineered human NSCs significantly migrated toward K562 CML cells more than human normal lung cells. In addition, the important chemoattractant factors involved in the specific migration ability of stem cells were found to be expressed in K562 CML cells. In a xenograft mouse model, NSC treatments via subcutaneous and intravenous injections resulted in significant inhibitions of tumor mass growth and extended survival dates of the mice. Taken together, these results suggest that gene therapy using genetically engineered stem cells expressing CD and IFN-β may be effective for treating CML in these mouse models.