A Chemo-enzymatically Linked Bispecific Antibody Retargets T Cells to a Sialylated Epitope on CD43 in Acute Myeloid Leukemia.

Acute myeloid leukemia (AML) is a high-risk disease with a poor prognosis, particularly in elderly patients. Because current AML treatment relies primarily on untargeted therapies with severe side effects that limit patient eligibility, identification of novel therapeutic AML targets is highly desired. We recently described AT1413, an antibody produced by donor B cells of a patient with AML cured after allogeneic hematopoietic stem cell transplantation. AT1413 binds CD43s, a unique sialylated epitope on CD43, which is weakly expressed on normal myeloid cells and overexpressed on AML cells. Because of its selectivity for AML cells, we considered CD43s as a target for a bispecific T-cell-engaging antibody (bTCE) and generated a bTCE by coupling AT1413 to two T-cell-targeting fragments using chemo-enzymatic linkage. In vitro, AT1413 bTCE efficiently induced T-cell-mediated cytotoxicity toward different AML cell lines and patient-derived AML blasts, whereas endothelial cells with low binding capacity for AT1413 remained unaffected. In the presence of AML cells, AT1413 bTCE induced upregulation of T-cell activation markers, cytokine release, and T-cell proliferation. AT1413 bTCE was also effective in vivo. Mice either coinjected with human peripheral blood mononuclear cells or engrafted with human hematopoietic stem cells [human immune system (HIS) mice] were inoculated with an AML cell line or patient-derived primary AML blasts. AT1413 bTCE treatment strongly inhibited tumor growth and, in HIS mice, had minimal effects on normal human hematopoietic cells. Taken together, our results indicate that CD43s is a promising target for T-cell-engaging antibodies and that AT1413 holds therapeutic potential in a bTCE-format. SIGNIFICANCE: These findings offer preclinical evidence for the therapeutic potential of a bTCE antibody that targets a sialylated epitope on CD43 in AML.

BOB.1 controls memory B-cell fate in the germinal center reaction.

During T cell-dependent (TD) germinal center (GC) responses, naïve B cells are instructed to differentiate towards GC B cells (GCBC), high-affinity long-lived plasma cells (LLPC) or memory B cells (Bmem). Alterations in the B cell-fate choice could contribute to immune dysregulation leading to the loss of self-tolerance and the initiation of autoimmune disease. Here we show that mRNA levels of the transcription regulator BOB.1 are increased in the lymph node compartment of patients with rheumatoid arthritis (RA), a prototypical autoimmune disease caused by the loss of immunological tolerance. Investigating to what extent levels of BOB.1 impact B cells during TD immune responses we found that BOB.1 has a crucial role in determining the B cell-fate decision. High BOB.1 levels promote the generation of cells with phenotypic and functional characteristics of Bmem. Mechanistically, overexpression of BOB.1 drives ABF1 and suppresses BCL6, favouring Bmem over LLPC or recycling GCBC. Low levels of BOB.1 are sufficient for LLPC but not for Bmem differentiation. Our findings demonstrate a novel role for BOB.1 in B cells during TD GC responses and suggest that its dysregulation may contribute to the pathogenesis of RA by disturbing the B cell-fate determination.

Establishment of Systems to Enable Isolation of Porcine Monoclonal Antibodies Broadly Neutralizing the Porcine Reproductive and Respiratory Syndrome Virus.

The rapid evolution of porcine reproductive and respiratory syndrome viruses (PRRSV) poses a major challenge to effective disease control since available vaccines show variable efficacy against divergent strains. Knowledge of the antigenic targets of virus-neutralizing antibodies that confer protection against heterologous PRRSV strains would be a catalyst for the development of next-generation vaccines. Key to discovering these epitopes is the isolation of neutralizing monoclonal antibodies (mAbs) from immune pigs. To address this need, we sought to establish systems to enable the isolation of PRRSV neutralizing porcine mAbs. We experimentally produced a cohort of immune pigs by sequential challenge infection with four heterologous PRRSV strains spanning PRRSV-1 subtypes and PRRSV species. Whilst priming with PRRSV-1 subtype 1 did not confer full protection against a subsequent infection with a PRRSV-1 subtype 3 strain, animals were protected against a subsequent PRRSV-2 infection. The infection protocol resulted in high serum neutralizing antibody titers against PRRSV-1 Olot/91 and significant neutralization of heterologous PRRSV-1/-2 strains. Enriched memory B cells isolated at the termination of the study were genetically programmed by transduction with a retroviral vector expressing the Bcl-6 transcription factor and the anti-apoptotic Bcl-xL protein, a technology we demonstrated efficiently converts porcine memory B cells into proliferating antibody-secreting cells. Pools of transduced memory B cells were cultured and supernatants containing PRRSV-specific antibodies identified by flow cytometric staining of infected MARC-145 cells and in vitro neutralization of PRRSV-1. Collectively, these data suggest that this experimental system may be further exploited to produce a panel of PRRSV-specific mAbs, which will contribute both to our understanding of the antibody response to PRRSV and allow epitopes to be resolved that may ultimately guide the design of immunogens to induce cross-protective immunity.

Cross-genotype AR3-specific neutralizing antibodies confer long-term protection in injecting drug users after HCV clearance.

BACKGROUND & AIMS: In order to design an effective vaccine against hepatitis C virus (HCV) infection, it is necessary to understand immune protection. A number of broadly reactive neutralizing antibodies have been isolated from B cells of HCV-infected patients. However, it remains unclear whether B cells producing such antibodies contribute to HCV clearance and long-term immune protection against HCV. METHODS: We analysed the B cell repertoire of 13 injecting drug users from the Amsterdam Cohort Study, who were followed up for a median of 17.5 years after primary infection. Individuals were classified into 2 groups based on the outcome of HCV infection: 5 who became chronically infected either after primary infection or after reinfection, and 8 who were HCV RNA negative following spontaneous clearance of ≥1 HCV infection(s). From each individual, 10,000 CD27+IgG+B cells, collected 0.75 year after HCV infection, were cultured to characterize the antibody repertoire. RESULTS: Using a multiplex flow cytometry-based assay to study the antibody binding to E1E2 from genotype 1 to 6, we found that a high frequency of cross-genotype antibodies was associated with spontaneous clearance of 1 or multiple infections (p = 0.03). Epitope specificity of these cross-genotype antibodies was determined by alanine mutant scanning in 4 individuals who were HCV RNA negative following spontaneous clearance of 1 or multiple infections. Interestingly, the cross-genotype antibodies were mainly antigenic region 3 (AR3)-specific and showed cross-neutralizing activity against HCV. In addition to AR3 antibodies, 3 individuals developed antibodies recognizing antigenic region 4, of which 1 monoclonal antibody showed cross-neutralizing capacity. CONCLUSIONS: Together, these data suggest that a strong B cell response producing cross-genotype and neutralizing antibodies, especially targeting AR3, contributes to HCV clearance and long-term immune protection against HCV. LAY SUMMARY: Although effective treatments against hepatitis C virus (HCV) are available, 500,000 people die from liver disease caused by HCV each year and approximately 1.75 million people are newly infected. This could be prevented by a vaccine. To design a vaccine against HCV, more insight into the role of antibodies in the protection against HCV infection is needed. In a cohort of injecting drug users, we found that antibodies interfering with virus cell entry, and recognizing multiple HCV genotypes, conferred long-term protection against chronic HCV infection.

Generation and Characterization of Anti-Citrullinated Protein Antibody-Producing B Cell Clones From Rheumatoid Arthritis Patients.

OBJECTIVE: Anti-citrullinated protein antibodies (ACPAs) are a hallmark of rheumatoid arthritis (RA). Aside from autoantibody production, the function of autoantigen-specific B cells remains poorly understood in the context of this disease. This study set out to elucidate autoantigen-specific B cell functions through the isolation and immortalization of unique citrullinated protein/peptide (CP)-reactive B cell clones from RA patients. METHODS: B cell clones from either the blood or synovial fluid of cyclic citrullinated peptide 2 (CCP2) antibody-positive RA patients were immortalized by genetic reprogramming with Bcl-6 and Bcl-xL. Enzyme-linked immunosorbent assay and flow cytometry were used to identify CCP2-reactive clones and to further characterize surface marker and cytokine expression as well as B cell receptor signaling competence. Global gene expression profiles were interrogated by RNA sequencing. RESULTS: Three unique CP-reactive memory B cell clones were generated from the blood or synovial fluid of 2 RA patients. CP-reactive memory B cells did not appear to be broadly cross-reactive, but rather had a fairly restricted epitope recognition profile. These clones were able to secrete both pro- and antiinflammatory cytokines and had a unique surface profile of costimulatory molecules and receptors, including CD40 and C5a receptor type 1, when compared to non-CP-reactive clones from the same patient. In addition, CP-reactive clones bound citrullinated protein, but not native protein, and could mobilize calcium in response to antigen binding. CONCLUSION: CP-reactive memory B cells comprise a rare, seemingly oligoclonal population with restricted epitope specificity and distinct phenotypic and molecular characteristics suggestive of antigen-presenting cells. Cloning by genetic reprogramming opens new avenues to study the function of autoreactive memory B cells, especially in terms of antigen processing, presentation, and subsequent T cell polarization.

AML-specific cytotoxic antibodies in patients with durable graft-versus-leukemia responses.

Most patients with acute myeloid leukemia (AML) can only be cured when allogeneic hematopoietic stem-cell transplantation induces a graft-versus-leukemia immune response (GVL). Although the role of T cells and natural killer cells in tumor immunology has been established, less is known about the contribution of B cells. From B cells of high-risk patients with AML with potent and lasting GVL responses, we isolated monoclonal antibodies directed against antigens expressed on the cell surface of AML cells but not on normal hematopoietic and nonhematopoietic cells. A number of these donor-derived antibodies recognized the U5 snRNP200 complex, a component of the spliceosome that in normal cells is found in the cell. In AML however, the U5 snRNP200 complex is exposed on the cell membrane of leukemic blasts. U5 snRNP200 complex-specific antibodies induced death of AML cells in an Fc receptor-dependent way in the absence of cytotoxic leukocytes or complement. In an AML mouse model, treatment with U5 snRNP200 complex-specific antibodies led to significant tumor growth inhibition. Thus, donor-derived U5 snRNP200 complex-recognizing AML-specific antibodies may contribute to antitumor responses.

Multiplex flow cytometry-based assay to study the breadth of antibody responses against E1E2 glycoproteins of hepatitis C virus.

Hepatitis C virus (HCV) infection is a major global public health problem. Early induction of cross-reactive neutralizing antibodies during acute infection correlates with the spontaneous clearance of HCV. Understanding the antibody response in multiple subjects in large-scale studies would greatly benefit vaccine development. To determine the breadth of a polyclonal-serum antibody response, and or, the monoclonal antibodies against the different HCV E1E2 genotypes, we developed a quick and high throughput flow cytometry assay using fluorescent cell barcoding to distinguish cells transfected with different E1E2 sequences in a single measurement. HCV-specific antibodies recognizing conformational epitopes were tested for binding to cells transfected with E1E2 from six genotypes. In this assay, 1500 samples can be analyzed for specific binding to 6 different HCV E1E2 sequences within 8h. Plasma of HCV infected subjects were tested in our assay allowing us to determine the breadth of their antibody response. In summary, we developed a quick and high throughput assay to study the specificity of an antibody response against multiple HCV E1E2 sequences simultaneously. This assay can also be used to facilitate the discovery of novel antibodies, and because other flavi- and picornaviruses have similar intracellular assembly mechanisms, this approach can be used to study the antibody response against such viruses.

Patient-derived antibody recognizes a unique CD43 epitope expressed on all AML and has antileukemia activity in mice.

Immunotherapy has proven beneficial in many hematologic and nonhematologic malignancies, but immunotherapy for acute myeloid leukemia (AML) and myelodysplastic syndrome (MDS) is hampered by the lack of tumor-specific targets. We took advantage of the tumor-immunotherapeutic effect of allogeneic hematopoietic stem cell transplantation and searched the B-cell repertoire of a patient with a lasting and potent graft-versus-AML response for the presence of AML-specific antibodies. We identified an antibody, AT1413, that was of donor origin and that specifically recognizes a novel sialylated epitope on CD43 (CD43s). Strikingly, CD43s is expressed on all World Health Organization 2008 types of AML and MDS. AT1413 induced antibody-dependent cell-mediated cytotoxicity and complement-dependent cytotoxicity of AML cells in vitro. Of note, AT1413 was highly efficacious against AML cells in a humanized mouse model without affecting nonmalignant human myeloid cells, suggesting AT1413 has potential as a therapeutic antibody.

Hepatitis C virus Broadly Neutralizing Monoclonal Antibodies Isolated 25 Years after Spontaneous Clearance.

Hepatitis C virus (HCV) is world-wide a major cause of liver related morbidity and mortality. No vaccine is available to prevent HCV infection. To design an effective vaccine, understanding immunity against HCV is necessary. The memory B cell repertoire was characterized from an intravenous drug user who spontaneously cleared HCV infection 25 years ago. CD27+IgG+ memory B cells were immortalized using BCL6 and Bcl-xL. These immortalized B cells were used to study antibody-mediated immunity against the HCV E1E2 glycoproteins. Five E1E2 broadly reactive antibodies were isolated: 3 antibodies showed potent neutralization of genotype 1 to 4 using HCV pseudotyped particles, whereas the other 2 antibodies neutralized genotype 1, 2 and 3 or 1 and 2 only. All antibodies recognized non-linear epitopes on E2. Finally, except for antibody AT12-011, which recognized an epitope consisting of antigenic domain C /AR2 and AR5, all other four antibodies recognized epitope II and domain B. These data show that a subject, who spontaneously cleared HCV infection 25 years ago, still has circulating memory B cells that are able to secrete broadly neutralizing antibodies. Presence of such memory B cells strengthens the argument for undertaking the development of an HCV vaccine.

Multiple capsid-stabilizing interactions revealed in a high-resolution structure of an emerging picornavirus causing neonatal sepsis.

The poorly studied picornavirus, human parechovirus 3 (HPeV3) causes neonatal sepsis with no therapies available. Our 4.3-Å resolution structure of HPeV3 on its own and at 15 Å resolution in complex with human monoclonal antibody Fabs demonstrates the expected picornavirus capsid structure with three distinct features. First, 25% of the HPeV3 RNA genome in 60 sites is highly ordered as confirmed by asymmetric reconstruction, and interacts with conserved regions of the capsid proteins VP1 and VP3. Second, the VP0 N terminus stabilizes the capsid inner surface, in contrast to other picornaviruses where on expulsion as VP4, it forms an RNA translocation channel. Last, VP1's hydrophobic pocket, the binding site for the antipicornaviral drug, pleconaril, is blocked and thus inappropriate for antiviral development. Together, these results suggest a direction for development of neutralizing antibodies, antiviral drugs based on targeting the RNA-protein interactions and dissection of virus assembly on the basis of RNA nucleation.

Structural Basis of Human Parechovirus Neutralization by Human Monoclonal Antibodies.

UNLABELLED: Since it was first recognized in 2004 that human parechoviruses (HPeV) are a significant cause of central nervous system and neonatal sepsis, their clinical importance, primarily in children, has started to emerge. Intravenous immunoglobulin treatment is the only treatment available in such life-threatening cases and has given moderate success. Direct inhibition of parechovirus infection using monoclonal antibodies is a potential treatment. We have developed two neutralizing monoclonal antibodies against HPeV1 and HPeV2, namely, AM18 and AM28, which also cross-neutralize other viruses. Here, we present the mapping of their epitopes using peptide scanning, surface plasmon resonance, fluorescence-based thermal shift assays, electron cryomicroscopy, and image reconstruction. We determined by peptide scanning and surface plasmon resonance that AM18 recognizes a linear epitope motif including the arginine-glycine-aspartic acid on the C terminus of capsid protein VP1. This epitope is normally used by the virus to attach to host cell surface integrins during entry and is found in 3 other viruses that AM18 neutralizes. Therefore, AM18 is likely to cause virus neutralization by aggregation and by blocking integrin binding to the capsid. Further, we show by electron cryomicroscopy, three-dimensional reconstruction, and pseudoatomic model fitting that ordered RNA interacts with HPeV1 VP1 and VP3. AM28 recognizes quaternary epitopes on the capsid composed of VP0 and VP3 loops from neighboring pentamers, thereby increasing the RNA accessibility temperature for the virus-AM28 complex compared to the virus alone. Thus, inhibition of RNA uncoating probably contributes to neutralization by AM28. IMPORTANCE: Human parechoviruses can cause mild infections to severe diseases in young children, such as neonatal sepsis, encephalitis, and cardiomyopathy. Intravenous immunoglobulin treatment is the only treatment available in such life-threatening cases. In order to develop more targeted treatment, we have searched for human monoclonal antibodies that would neutralize human parechoviruses 1 and 2, associated with mild infections such as gastroenteritis and severe infections of the central nervous system, and thus allow safe treatment. In the current study, we show how two such promising antibodies interact with the virus, modeling the atomic interactions between the virus and the antibody to propose how neutralization occurs. Both antibodies can cause aggregation; in addition, one antibody interferes with the virus recognizing its target cell, while the other, recognizing only the whole virus, inhibits the genome uncoating and replication in the cell.

Bispecific antibody generated with sortase and click chemistry has broad antiinfluenza virus activity.

Bispecific antibodies have therapeutic potential by expanding the functions of conventional antibodies. Many different formats of bispecific antibodies have meanwhile been developed. Most are genetic modifications of the antibody backbone to facilitate incorporation of two different variable domains into a single molecule. Here, we present a bispecific format where we have fused two full-sized IgG antibodies via their C termini using sortase transpeptidation and click chemistry to create a covalently linked IgG antibody heterodimer. By linking two potent anti-influenza A antibodies together, we have generated a full antibody dimer with bispecific activity that retains the activity and stability of the two fusion partners.

A common solution to group 2 influenza virus neutralization.

The discovery and characterization of broadly neutralizing antibodies (bnAbs) against influenza viruses have raised hopes for the development of monoclonal antibody (mAb)-based immunotherapy and the design of universal influenza vaccines. Only one human bnAb (CR8020) specifically recognizing group 2 influenza A viruses has been previously characterized that binds to a highly conserved epitope at the base of the hemagglutinin (HA) stem and has neutralizing activity against H3, H7, and H10 viruses. Here, we report a second group 2 bnAb, CR8043, which was derived from a different germ-line gene encoding a highly divergent amino acid sequence. CR8043 has in vitro neutralizing activity against H3 and H10 viruses and protects mice against challenge with a lethal dose of H3N2 and H7N7 viruses. The crystal structure and EM reconstructions of the CR8043-H3 HA complex revealed that CR8043 binds to a site similar to the CR8020 epitope but uses an alternative angle of approach and a distinct set of interactions. The identification of another antibody against the group 2 stem epitope suggests that this conserved site of vulnerability has great potential for design of therapeutics and vaccines.

Genetic manipulation of B cells for the isolation of rare therapeutic antibodies from the human repertoire.

Antibody based therapies are increasingly applied to prevent and treat human disease. While the majority of antibodies currently on the market are chimeric or humanized antibodies from rodents, the focus has now shifted to the isolation and development of fully human antibodies. By retroviral transduction of B cell lymphoma-6 (BCL-6), which prevents terminal differentiation of B cells and, the anti-apoptotic gene B-cell lymphoma-extra large (Bcl-xL) into primary human B cells we efficiently immortalize antibody-producing B cells allowing the isolation of therapeutic antibodies. Selection of antigen-specific B cell clones was greatly facilitated because the transduced B cells retain surface immunoglobulin expression and secrete immunoglobulin into the culture supernatant. Surface immunoglobulin expression can be utilized to stain and isolate antigen specific B cell clones with labeled antigen. Immunoglobulins secreted in culture supernatant can directly be tested in functional assays to identify unique B cell clones. Here we describe the key features of our Bcl-6/Bcl-xL culture platform (AIMSelect).

Structure-based design of a fusion glycoprotein vaccine for respiratory syncytial virus.

Respiratory syncytial virus (RSV) is the leading cause of hospitalization for children under 5 years of age. We sought to engineer a viral antigen that provides greater protection than currently available vaccines and focused on antigenic site Ø, a metastable site specific to the prefusion state of the RSV fusion (F) glycoprotein, as this site is targeted by extremely potent RSV-neutralizing antibodies. Structure-based design yielded stabilized versions of RSV F that maintained antigenic site Ø when exposed to extremes of pH, osmolality, and temperature. Six RSV F crystal structures provided atomic-level data on how introduced cysteine residues and filled hydrophobic cavities improved stability. Immunization with site Ø-stabilized variants of RSV F in mice and macaques elicited levels of RSV-specific neutralizing activity many times the protective threshold.

Novel staphylococcal glycosyltransferases SdgA and SdgB mediate immunogenicity and protection of virulence-associated cell wall proteins.

Infection of host tissues by Staphylococcus aureus and S. epidermidis requires an unusual family of staphylococcal adhesive proteins that contain long stretches of serine-aspartate dipeptide-repeats (SDR). The prototype member of this family is clumping factor A (ClfA), a key virulence factor that mediates adhesion to host tissues by binding to extracellular matrix proteins such as fibrinogen. However, the biological siginificance of the SDR-domain and its implication for pathogenesis remain poorly understood. Here, we identified two novel bacterial glycosyltransferases, SdgA and SdgB, which modify all SDR-proteins in these two bacterial species. Genetic and biochemical data demonstrated that these two glycosyltransferases directly bind and covalently link N-acetylglucosamine (GlcNAc) moieties to the SDR-domain in a step-wise manner, with SdgB appending the sugar residues proximal to the target Ser-Asp repeats, followed by additional modification by SdgA. GlcNAc-modification of SDR-proteins by SdgB creates an immunodominant epitope for highly opsonic human antibodies, which represent up to 1% of total human IgG. Deletion of these glycosyltransferases renders SDR-proteins vulnerable to proteolysis by human neutrophil-derived cathepsin G. Thus, SdgA and SdgB glycosylate staphylococcal SDR-proteins, which protects them against host proteolytic activity, and yet generates major eptopes for the human anti-staphylococcal antibody response, which may represent an ongoing competition between host and pathogen.

Functional CD47/signal regulatory protein alpha (SIRP(alpha)) interaction is required for optimal human T- and natural killer- (NK) cell homeostasis in vivo.

The homeostatic control mechanisms regulating human leukocyte numbers are poorly understood. Here, we assessed the role of phagocytes in this process using human immune system (HIS) BALB/c Rag2(-/-)IL-2Rγc(-/-) mice in which human leukocytes are generated from transplanted hematopoietic progenitor cells. Interactions between signal regulatory protein alpha (SIRPα; expressed on phagocytes) and CD47 (expressed on hematopoietic cells) negatively regulate phagocyte activity of macrophages and other phagocytic cells. We previously showed that B cells develop and survive robustly in HIS mice, whereas T and natural killer (NK) cells survive poorly. Because human CD47 does not interact with BALB/c mouse SIRPα, we introduced functional CD47/SIRPα interactions in HIS mice by transducing mouse CD47 into human progenitor cells. Here, we show that this procedure resulted in a dramatic and selective improvement of progenitor cell engraftment and human T- and NK-cell homeostasis in HIS mouse peripheral lymphoid organs. The amount of engrafted human B cells also increased but much less than that of T and NK cells, and total plasma IgM and IgG concentrations increased 68- and 35-fold, respectively. Whereas T cells exhibit an activated/memory phenotype in the absence of functional CD47/SIRPα interactions, human T cells accumulated as CD4(+) or CD8(+) single-positive, naive, resting T cells in the presence of functional CD47/SIRPα interactions. Thus, in addition to signals mediated by T cell receptor (TCR)/MHC and/or IL/IL receptor interactions, sensing of cell surface CD47 expression by phagocyte SIRPα is a critical determinant of T- and NK-cell homeostasis under steady-state conditions in vivo.

Synergy between IL-15 and Id2 promotes the expansion of human NK progenitor cells, which can be counteracted by the E protein HEB required to drive T cell development.

The cytokine IL-15 and the inhibitor of DNA binding (Id)2, which negatively regulates the activity of basic helix-loop-helix transcription factors, have been shown to play key roles in NK cell development. Consistent with this, exogenous IL-15 added to human thymic progenitor cells stimulated their development into NK cells at the expense of T cells both in fetal thymic organ culture and in coculture with stromal cells expressing the Notch ligand Delta-like 1. Overexpression of Id2 in thymic progenitor cells stimulated NK cell development and blocked T cell development. This, in part, is attributed to inhibition of the transcriptional activity of the E protein HEB, which we show in this study is the only E protein that enhanced T cell development. Notably, Id2 increased a pool of lineage CD1a-CD5+ progenitor cells that in synergy with IL-15 furthered expansion and differentiation into NK cells. Taken together, our findings point to a dualistic function of Id2 in controlling T/NK cell lineage decisions; T cell development is impaired by Id2, most likely by sequestering HEB, whereas NK cell development is promoted by increasing a pool of CD1a-CD5+ NK cell progenitors, which together with IL-15 differentiate into mature NK cells.