Behavior of transplanted bone marrow-derived GFP mesenchymal cells in osteochondral defect as a simulation of autologous transplantation.

To elucidate the behavior of autologously transplanted mesenchymal cells in osteochondral defects, we followed transplanted cells using green fluorescent protein (GFP) transgenic rats, in which all cells express GFP signals in their cytoplasm and nuclei as transplantation donors. Bone marrow-derived mesenchymal cells, which contain mesenchymal stem cells (MSCs), were obtained from transgenic rats. Then, dense mesenchymal cell masses created by hanging-drop culture were transplanted and fixed with fibrin glue into osteochondral defects of wild-type rats. At 24 weeks after surgery, the defects were repaired with hyaline-like cartilage and subchondral bone. GFP positive cells, indicating transplanted mesenchymal-derived cells, were observed in the regenerated tissues for 24 weeks although GFP positive cells decreased in number with time. Because GFP causes no immunological rejection and requires no chemicals for visualization, transplantation between transgenic and wild-type rats can be regarded as a simulation of autologous transplantation, and the survivability of transplanted cells are able to be followed easily and reliably. Thus, the behavior of transplanted mesenchymal cells was able to be elucidated in vivo by this strategy, and the results could be essential in future tissue engineering for the regeneration of osteochondral defects with original hyaline cartilage and subchondral bone.     

A subpopulation of bone marrow cells depleted by a novel antibody, anti-Liv8, is useful for cell therapy to repair damaged liver

We previously reported a new in vivo model named as "GFP/CCl(4) model" for monitoring the transdifferentiation of green fluorescent protein (GFP) positive bone marrow cell (BMC) into albumin-positive hepatocyte under the specific "niche" made by CCl(4) induced persistent liver damage, but the subpopulation which BMCs transdifferentiate into hepatocytes remains unknown. Here we developed a new monoclonal antibody, anti-Liv8, using mouse E 11.5 fetal liver as an antigen. Anti-Liv8 recognized both hematopoietic progenitor cells in fetal liver at E 11.5 and CD45-positive hematopoietic cells in adult bone marrow. We separated Liv8-positive and Liv8-negative cells and then transplanted these cells into a continuous liver damaged model. At 4 weeks after BMC transplantation, more efficient repopulation and transdifferentiation of BMC into hepatocytes were seen with Liv8-negative cells. These findings suggest that the subpopulation of Liv8-negative cells includes useful cells to perform cell therapy on repair damaged liver.     

Introduction of the green fluorescent protein gene into hematopoietic stem cells results in prolonged discrepancy of in vivo transduction levels between bone marrow progenitors and peripheral blood cells in nonhuman primates

Background

The green fluorescent protein (GFP) has proven a useful marker in retroviral gene transfer studies targeting hematopoietic stem cells (HSCs) in mice. However, several investigators have reported very low in vivo peripheral blood marking levels in nonhuman primates after transplantation of HSCs transduced with the GFP gene. We retrovirally marked cynomolgus monkey HSCs with the GFP gene, and tracked in vivo marking levels within both bone marrow progenitor cells and mature peripheral blood cells following autologous transplantation after myeloablative conditioning.

Methods

Bone marrow cells were harvested from three cynomolgus macaques and enriched for the primitive fraction by CD34 selection. CD34⁺ cells were transduced with one of three retroviral vectors all expressing the GFP gene and were infused after myeloablative total body irradiation (500 cGy × 2). Following transplantation, proviral levels and fluorescence were monitored among clonogenic bone marrow progenitors and mature peripheral blood cells.

Results

Although 13–37% of transduced cells contained the GFP provirus and 11–13% fluoresced ex vivo, both provirus and fluorescence became almost undetectable in the peripheral blood within several months after transplantation regardless of the vectors used. However, on sampling of bone marrow at multiple time points, significant fractions (5–10%) of clonogenic progenitors contained the provirus and fluoresced ex vivo reflecting a significant discrepancy between GFP gene marking levels within bone marrow cells and their mature peripheral blood progeny. The discrepancy (at least one log) persisted for more than 1 year after transplantation. Since no cytotoxic T lymphocytes against GFP were detected in the animals, an immune response against GFP is an unlikely explanation for the low levels of transduced peripheral blood cells. Administration of granulocyte colony stimulating factor and stem cell factor resulted in mobilization of transduced bone marrow cells detectable as mature granulocyte progeny which expressed the GFP gene, suggesting that transduced progenitor cells in bone marrow could be mobilized into the peripheral blood and differentiated into granulocytes.

Conclusions

Low levels of GFP‐transduced mature cells in the peripheral blood of nonhuman primates may reflect a block to differentiation associated with GFP; this block can be overcome in part by nonphysiological cytokine treatment ex vivo and in vivo. Copyright © 2002 John Wiley & Sons, Ltd.

     

Bioimaging for the monitoring of the in vivo distribution of infused mesenchymal stem cells in a mouse model of the graft-versus-host reaction

Cell therapy using MSCs (mesenchymal stem cells) might be effective treatment for refractory GVHD (graft-versus-host disease). However, the fate and distribution of MSCs after transplantation remains unclear. In this study, an animal model was developed to monitor the dynamic distribution of MSCs in mice with GVHD. A GVHD mouse model was established by transplanting C57BL/6 donor bone marrow cells and C57BL/6 EGFP (enhanced green fluorescent protein) splenocytes into lethally irradiated BALB/c nude recipient mice. Donor MSCs were obtained from MHC-identical C57BL/6 RFP (red fluorescent protein) mice and infused into the recipient mice on the same transplantation day. In vivo movement of the donor splenocytes (EGFP) and MSCs (RFP) were evaluated by measuring the biofluorescence (IVIS-Xenogen system). Donor splenocytes and MSCs reached the lungs first, and then the gastrointestinal tract, lymph nodes and skin, in that order; the transit time and localization site of these cells were very similar. In the recipient mouse with GVHD, the number of detectable cells declined with time, as assessed by biofluorescence imaging and confirmed by RT (real-time)-PCR. This bioimaging system might be useful for preclinical testing and the design of therapeutic strategies for monitoring the dynamic distribution of MSCs with GVHD.     

Application of dedifferentiated fat cells for periodontal tissue regeneration

Periodontal diseases result from inflammation by bacterial infection in plaques, leading to tooth loss. However, regenerative approaches with periodontal tissue regeneration by guided tissue regeneration and enamel matrix derivative are not yet well established. Tissue regeneration requires three factors: cells, scaffold, and growth factors. Dedifferentiated fat cells (DFATs) are pluripotent with the same differentiation capacities as mesenchymal stem cells (MSCs). Access to MSCs is limited, whereas donor cells for DFATs are abundant in adipose tissues and can be non-invasively obtained. Therefore, we tested DFATs as a new source for periodontal tissue regeneration in an experimental periodontal tissue loss model in rats by transplanting DFATs on an atelocollagen scaffold using DFATs isolated from Sprague–Dawley (SD) rats expressing green fluorescent protein (GFP). GFP–DFAT cells were transplanted on the palatal side of the upper left first molar in SD rats and detected by H&E staining, GFP, and proliferating cell nuclear antigen (PCNA) expression. DFAT differentiation was also evaluated in three-dimensional cultures. GFP positive cells were detected in the regenerated tissue by the DFATs/scaffold mixture at 4 weeks after transplantation, and PCNA-positive cells were significantly increased in the periodontal ligament along the new bone in the DFATs/scaffold group more than in the scaffold-only group, suggesting that DFATs differentiate in the same manner as MSCs and regenerate in the defective areas. Consistent with previous reports, DFATs differentiation was slower than that with stem cells. The present study demonstrates that DFATs are pluripotent and an effective new source of cells for periodontal tissue regeneration.     

Visualization of GFP-expressing tumors and metastasis in vivo

We have developed mouse models of metastatic cancer with genetically fluorescent tumors that can be imaged in fresh tissue, in situ, as well as externally. To achieve this capability, we have transduced the green fluorescent protein (GFP) gene, cloned from the bioluminescent jellyfish Aequorea victoria, into a series of human and rodent cancer cell lines that were selected in vitro to stably express GFP in vivo after transplantation to metastatic rodent models. Techniques were also developed for transduction of tumors by GFP in vivo. With this fluorescent tool, we detected and visualized for the first time tumors and metastasis in fresh viable tissue or in situ in host organs down to the single-cell level. GFP tumors on the colon, prostate, breast, brain, liver, lymph nodes, lung, pancreas, bone, and other organs can also be visualized externally, transcutaneously by quantitative whole-body fluorescence optical imaging. Real-time tumor and metastatic growth and angiogenesis and inhibition by representative drugs can be imaged and quantified for rapid antitumor, antimetastatic, and antiangiogenesis drug screening. The GFP-transfected tumor cells enabled a fundamental advance in the visualization of tumor growth and metastasis in real time in vivo.     

Neogenesis of cerebellar Purkinje neurons from gene-marked bone marrow cells in vivo.

The versatility of stem cells has only recently been fully recognized. There is evidence that upon adoptive bone marrow (BM) transplantation (BMT), donor-derived cells can give rise to neuronal phenotypes in the brains of recipient mice. Yet only few cells with the characteristic shape of neurons were detected 1-6 mo post-BMT using transgenic or newborn mutant mice. To evaluate the potential of BM to generate mature neurons in adult C57BL/6 mice, we transferred the enhanced green fluorescent protein (GFP) gene into BM cells using a murine stem cell virus-based retroviral vector. Stable and high level long-term GFP expression was observed in mice transplanted with the transduced BM. Engraftment of GFP-expressing cells in the brain was monitored by intravital microscopy. In a long-term follow up of 15 mo post-BMT, fully developed Purkinje neurons were found to express GFP in both cerebellar hemispheres and in all chimeric mice. GFP-positive Purkinje cells were also detected in BM chimeras from transgenic mice that ubiquitously express GFP. Based on morphologic criteria and the expression of glutamic acid decarboxylase, the newly generated Purkinje cells were functional.     

Immunotherapy of established tumors using bone marrow transplantation with antigen gene--modified hematopoietic stem cells

A major focus of cancer immunotherapy is to develop strategies to induce T-cell responses through presentation of tumor antigens by dendritic cells (DCs). Current vaccines are limited in their ability to efficiently transfer antigens to DCs in vivo. Ex vivo-generated DCs can be efficiently loaded with antigen but after reinjection, few DCs traffic to secondary lymphoid organs, the critical sites for antigen presentation. To enhance efficiency and durability of antigen presentation by DCs, we transduced hematopoietic stem-progenitor cells (HSCs) with a model tumor antigen and then transplanted the gene-modified cells into irradiated recipient mice, which resulted in efficient expression of the transgene in a large proportion of donor derived DCs in lymphoid organs. The combination of bone marrow transplantation (BMT) using transduced HSCs, systemic agents that generate and activate DCs, and mature T-cell infusion resulted in substantial expansion and activation of antigen-specific T cells. This tripartite strategy provided potent antigen-specific immunotherapy for an aggressive established tumor.     

Extrahepatic cells contribute to the progenitor/stem cell response following reduced-size liver transplantation in mice

The extent to which extrahepatic cells participate in liver regeneration following transplantation is not known. Either full-size or reduced-size livers from wild-type mice were implanted into green fluorescent protein-positive (GFP(+)) transgenic recipient mice to determine whether regenerated liver contained host-derived GFP(+) hepatic cells. After reduced-size liver transplantation, GFP(+) cells were localized to the portal zone of the liver lobule. Interestingly, GFP(+) cells stained for CD117, a marker for progenitor cells, beginning 2 days after transplantation. A significant number of GFP(+) CD117(+) cells were identified in donor livers after 28 days. GFP(+) cells comprised nearly 9% of the donor liver 28 days after reduced-size liver transplant. Moreover, GFP(+) cells also expressed the hepatic progenitor cell marker A6 and novel marker hepatic-specific antigen (HSA), as well as stem cell antigen-1 (Sca-1). Interestingly, some GFP(-) cells also were stained for CD117 and A6, suggesting that both extrahepatic and intrahepatic stem cells were present and may have contributed to the regenerative response under these conditions. Reduced-size liver transplantation using GFP(+) transgenic mice supports the hypothesis that recipient-derived progenitor cells are present and may contribute to liver regeneration following transplantation.     

A transgenic red fluorescent protein‐expressing nude mouse for color‐coded imaging of the tumor microenvironment

The tumor microenvironment (TME) is critical for tumor growth and progression. We have previously developed color‐coded imaging of the TME using a green fluorescent protein (GFP) transgenic nude mouse as a host. However, most donor sources of cell types appropriate for study in the TME are from mice expressing GFP. Therefore, a nude mouse expressing red fluorescent protein (RFP) would be an appropriate host for transplantation of GFP‐expressing stromal cells as well as double‐labeled cancer cells expressing GFP in the nucleus and RFP in the cytoplasm, thereby creating a three‐color imaging model of the TME. The RFP nude mouse was obtained by crossing non‐transgenic nude mice with the transgenic C57/B6 mouse in which the β‐actin promoter drives RFP (DsRed2) expression in essentially all tissues. In crosses between nu/nu RFP male mice and nu/+ RFP female mice, the embryos fluoresced red. Approximately 50% of the offspring of these mice were RFP nude mice. In the RFP nude mouse, the organs all brightly expressed RFP, including the heart, lungs, spleen, pancreas, esophagus, stomach, duodenum, the male and female reproductive systems; brain and spinal cord; and the circulatory system, including the heart, and major arteries and veins. The skinned skeleton highly expressed RFP. The bone marrow and spleen cells were also RFP positive. GFP‐expressing human cancer cell lines, including HCT‐116‐GFP colon cancer and MDA‐MB‐435‐GFP breast cancer were orthotopically transplanted to the transgenic RFP nude mice. These human tumors grew extensively in the transgenic RFP nude mouse. Dual‐color fluorescence imaging enabled visualization of human tumor–host interaction. The RFP nude mouse model should greatly expand our knowledge of the TME. J. Cell. Biochem. 106: 279–284, 2009. © 2008 Wiley‐Liss, Inc.     

Bone marrow origin of cells with capacity for homing and differentiation to esophageal squamous epithelium

Our goal was to determine whether esophageal progenitor cells could be isolated from adult mouse esophagus or bone marrow and shown to home to and proliferate in the irradiated esophagus of recipient mice. Esophageal progenitor cells were isolated from adult male C3H/HeNsd or C57BL/6J green fluorescent protein (GFP(+)) mice by a serial in vitro preplate technique or the technique of side population cell sorting. When injected intravenously (i.v.), these cells homed to the 30-Gy-irradiated esophagus of GFP(-) female recipient mice and formed donor-origin esophageal foci. GFP(+) whole murine bone marrow cells injected i.v. also formed donor-origin esophageal squamous cell foci and protected recipient GFP(-) mice from upper-body irradiation in a cell dose-dependent manner. Marrow chimeric GFP(-) mice reconstituted with GFP(+) cells showed migration of GFP(+) marrow cells to the esophagus after 30 Gy irradiation. Purified esophageal progenitor cells isolated from first-generation preplate cell recipients engrafted after i.v. injection to the esophagus of second-generation-irradiated recipient mice. These data establish that esophageal progenitor cells can home to the irradiated esophagus and show limited differentiation capacity to squamous epithelium.     

Xenotransplantation of Human Adipose-Derived Stem Cells in Zebrafish Embryos

Zebrafish is a widely used animal model with well-characterized background in developmental biology. The fate of human adipose-derived stem cells (ADSCs) after their xenotransplantation into the developing embryos of zebrafish is unknown. Therefore, human ADSCs were firstly isolated, and then transduced with lentiviral vector system carrying a green fluorescent protein (GFP) reporter gene, and followed by detection of their cell viability and the expression of cell surface antigens. These GFP-expressing human ADSCs were transplanted into the zebrafish embryos at 3.3–4.3 hour post-fertilization (hpf). Green fluorescent signal, the proliferation and differentiation of human ADSCs in recipient embryos were respectively examined using fluorescent microscopy and immunohistochemical staining. The results indicated that human ADSCs did not change their cell viability and the expression levels of cell surface antigens after GFP transduction. Microscopic examination demonstrated that green fluorescent signals of GFP expressed in the transplanted cells were observed in the embryos and larva fish at post-transplantation. The positive staining of Ki-67 revealed the survival and proliferation of human ADSCs in fish larvae after transplantation. The expression of CD105 was observable in the xenotransplanted ADSCs, but CD31 expression was undetectable. Therefore, our results indicate that human ADSCs xenotransplanted in the zebrafish embryos not only can survive and proliferate at across-species circumstance, but also seem to maintain their undifferentiation status in a short term. This xenograft model of zebrafish embryos may provide a promising and useful technical platform for the investigation of biology and physiology of stem cells in vivo.     

Busulfan as a Myelosuppressive Agent for Generating Stable High-level Bone Marrow Chimerism in Mice

Bone marrow transplantation (BMT) is often used to replace the bone marrow (BM) compartment of recipient mice with BM cells expressing a distinct biomarker isolated from donor mice. This technique allows for identification of donor-derived hematopoietic cells within the recipient mice, and can be used to isolate and characterize donor cells using various biochemical techniques. BMT typically relies on myeloablative conditioning with total body irradiation to generate niche space within the BM compartment of recipient mice for donor cell engraftment. The protocol we describe here uses myelosuppressive conditioning with the chemotherapeutic agent busulfan. Unlike irradiation, which requires the use of specialized facilities, busulfan conditioning is performed using intraperitoneal injections of 20 mg/kg busulfan until a total dose of 60-100 mg/kg has been administered. Moreover, myeloablative irradiation can have toxic side effects and requires successful engraftment of donor cells for survival of recipient mice. In contrast, busulfan conditioning using these doses is generally well tolerated and mice survive without donor cell support. Donor BM cells are isolated from the femurs and tibiae of mice ubiquitously expressing green fluorescent protein (GFP), and injected into the lateral tail vein of conditioned recipient mice. BM chimerism is estimated by quantifying the number of GFP+ cells within the peripheral blood following BMT. Levels of chimerism >80% are typically observed in the peripheral blood 3-4 weeks post-transplant and remain established for at least 1 year. As with irradiation, conditioning with busulfan and BMT allows for the accumulation of donor BM-derived cells within the central nervous system (CNS), particularly in mouse models of neurodegeneration. This busulfan-mediated CNS accumulation may be more physiological than total body irradiation, as the busulfan treatment is less toxic and CNS inflammation appears to be less extensive. We hypothesize that these cells can be genetically engineered to deliver therapeutics to the CNS.     

The role of bone marrow-derived cells in bone fracture repair in a green fluorescent protein chimeric mouse model

We investigated the role of bone marrow cells in bone fracture repair using green fluorescent protein (GFP) chimeric model mice. First, the chimeric model mice were created: bone marrow cells from GFP-transgenic C57BL/6 mice were injected into the tail veins of recipient wild-type C57BL/6 mice that had been irradiated with a lethal dose of 10Gy from a cesium source. Next, bone fracture models were created from these mice: closed transverse fractures of the left femur were produced using a specially designed device. One, three, and five weeks later, fracture lesions were extirpated for histological and immunohistochemical analyses. In the specimens collected 3 and 5 weeks after operation, we confirmed calluses showing intramembranous ossification peripheral to the fracture site. The calluses consisted of GFP- and osteocalcin-positive cells at the same site, although the femur consisted of only osteocalcin-positive cells. We suggest that bone marrow cells migrated outside of the bone marrow and differentiated into osteoblasts to make up the calluses.     

同种异体宫内移植小鼠嵌合模型的建立

干细胞宫内移植是一种很有前途的产前治疗方式。为深入研究干细胞移植后的细胞行为,采用宫内移植的方法建立同种异体的嵌合小鼠模型。将雄鼠骨髓单核细胞宫内注射到胎鼠腹腔,在受体鼠出生后检测雌性受体鼠外周血细胞。应用PCR检测外周血是否存在雄性鼠的DNA,并采用定量PCR技术确定其嵌合量;同时用荧光原位杂交（FISH）技术直观观察外周血中雄性来源的细胞。结果表明：共获得4只阳性外周血嵌合小鼠,其中3只稳定嵌合达到6个月以上。应用宫内移植成功建立了外周血中存在异源细胞的小鼠嵌合模型。     

Expansion of murine spermatogonial stem cells through serial transplantation

Mammalian male germ cells might be generally thought to have infinite proliferative potential based on their life-long production of huge numbers of sperm. However, there has been little substantial evidence that supports this assumption. In the present study, we performed serial transplantation of spermatogonial stem cells to investigate if they expand by self-renewing division following transplantation. The transgenic mouse carrying the Green fluorescent protein gene was used as the donor cell source that facilitated identification and recollection of colonized donor germ cells in the recipient testes. The established colonies of germ cells in the recipient testes were collected and transplanted to new recipients. This serial transplantation of spermatogonial stem cells repopulated the recipient testes, which were successfully performed sequentially up to four times from one recipient to the next. The incubation periods between two sequential transplantations ranged from 55 to 373 days. During these passages, the spermatogonial stem cells showed constant activity to form spermatogenic colonies in the recipient testis. They continued to increase in number for more than a year following transplantation. Colonization efficiency of spermatogonial stem cells was determined to be 4.25% by using Sl/Sl(d) mice as recipients that propagated only undifferentiated type A spermatogonia in their testes. Based on the colonization efficiency, one colony-forming activity was assessed to equate to about 20 spermatogonial stem cells. The spermatogonial stem cells were estimated to expand over 50-fold in 100 days in this experiment.     

Differentiating Functional Roles of Gene Expression from Immune and Non-immune Cells in Mouse Colitis by Bone Marrow Transplantation

To understand the role of a gene in the development of colitis, we compared the responses of wild-type mice and gene-of-interest deficient knockout mice to colitis. If the gene-of-interest is expressed in both bone marrow derived cells and non-bone marrow derived cells of the host; however, it is possible to differentiate the role of a gene of interest in bone marrow derived cells and non- bone marrow derived cells by bone marrow transplantation technique. To change the bone marrow derived cell genotype of mice, the original bone marrow of recipient mice were destroyed by irradiation and then replaced by new donor bone marrow of different genotype. When wild-type mice donor bone marrow was transplanted to knockout mice, we could generate knockout mice with wild-type gene expression in bone marrow derived cells. Alternatively, when knockout mice donor bone marrow was transplanted to wild-type recipient mice, wild-type mice without gene-of-interest expressing from bone marrow derived cells were produced. However, bone marrow transplantation may not be 100% complete. Therefore, we utilized cluster of differentiation (CD) molecules (CD45.1 and CD45.2) as markers of donor and recipient cells to track the proportion of donor bone marrow derived cells in recipient mice and success of bone marrow transplantation. Wild-type mice with CD45.1 genotype and knockout mice with CD45.2 genotype were used. After irradiation of recipient mice, the donor bone marrow cells of different genotypes were infused into the recipient mice. When the new bone marrow regenerated to take over its immunity, the mice were challenged by chemical agent (dextran sodium sulfate, DSS 5%) to induce colitis. Here we also showed the method to induce colitis in mice and evaluate the role of the gene of interest expressed from bone-marrow derived cells. If the gene-of-interest from the bone derived cells plays an important role in the development of the disease (such as colitis), the phenotype of the recipient mice with bone marrow transplantation can be significantly altered. At the end of colitis experiments, the bone marrow derived cells in blood and bone marrow were labeled with antibodies against CD45.1 and CD45.2 and their quantitative ratio of existence could be used to evaluate the success of bone marrow transplantation by flow cytometry. Successful bone marrow transplantation should show a vast majority of donor genotype (in term of CD molecule marker) over recipient genotype in both the bone marrow and blood of recipient mice.     

Determinants of antileukemia effects of allogeneic NK cells

In HLA-nonidentical bone marrow transplantation, we studied the characteristics of donor NK cells, recipient leukemia cells, and the cytokine environment that predict the antileukemia effects of allogeneic NK cells. We found that the risk of relapse in pediatric patients with hematologic malignancies was best predicted by a model taking into consideration the presence of inhibitory killer cell Ig-like receptors (KIRs) on the donor's NK cells and the absence of corresponding KIR ligand in the recipient's HLA repertoire (a receptor-ligand model). The risk of relapse was prognosticated less precisely by the Perugia donor-recipient KIR ligand-ligand mismatch model or by a natural cytotoxicity model. In contrast to the ligand-ligand model, we found that the new receptor-ligand model was accurate when analysis was applied to patients with lymphoid malignancy. These findings corroborate our observations that the recipient's KIR repertoire, which was derived from highly purified, HLA-disparate CD34+ cells, resumed a donor-specific pattern within 3 mo of transplantation, but did not correlate evidently with the donor or recipient ligand repertoire. In an in vitro assay and an in vivo mouse model, human NK cell cytotoxicity toward human leukemia cells with 11q23 chromosomal rearrangement increased with the number of receptor-ligand mismatch pairs or prestimulation with IL-12 and IL-18. These findings provide new insights into the determinants of antileukemia effects of allogeneic NK cells and therapeutic strategies.