Cytolytic Activity Cells Acquire Functional Receptors and Human Embryonic Stem Cell-Derived NK

Human Embryonic Stem Cell-Derived NK Cells Acquire Functional Receptors and Cytolytic Activity

This information is current as Petter S. Woll, Colin H. Martin, Jeffrey S. Miller and Dan S. of September 23, 2021. Kaufman J Immunol 2005; 175:5095-5103; ; doi: 10.4049/jimmunol.175.8.5095 http://www.jimmunol.org/content/175/8/5095 Downloaded from

References This article cites 57 articles, 29 of which you can access for free at: http://www.jimmunol.org/content/175/8/5095.full#ref-list-1 http://www.jimmunol.org/ Why The JI? Submit online.

• Rapid Reviews! 30 days* from submission to initial decision

• No Triage! Every submission reviewed by practicing scientists

• Fast Publication! 4 weeks from acceptance to publication

by guest on September 23, 2021 *average

Subscription Information about subscribing to The Journal of Immunology is online at: http://jimmunol.org/subscription Permissions Submit copyright permission requests at: http://www.aai.org/About/Publications/JI/copyright.html Email Alerts Receive free email-alerts when new articles cite this article. Sign up at: http://jimmunol.org/alerts

The Journal of Immunology is published twice each month by The American Association of Immunologists, Inc., 1451 Rockville Pike, Suite 650, Rockville, MD 20852 Copyright © 2005 by The American Association of Immunologists All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606. The Journal of Immunology

Human Embryonic Stem Cell-Derived NK Cells Acquire Functional Receptors and Cytolytic Activity1

Petter S. Woll,* Colin H. Martin,* Jeffrey S. Miller,† and Dan S. Kaufman2*

Human embryonic stem cells (hESCs) provide a unique resource to analyze early stages of human hematopoiesis. However, little is known about the ability to use hESCs to evaluate lymphocyte development. In the present study, we use a two-step culture method to demonstrate efﬁcient generation of functional NK cells from hESCs. The CD56؉CD45؉ hESC-derived lymphocytes express inhibitory and activating receptors typical of mature NK cells, including killer cell Ig-like receptors, natural cytotoxicity receptors, and CD16. Limiting dilution analysis suggests that these cells can be produced from hESC-derived hemopoietic progenitors at a clonal frequency similar to CD34؉ cells isolated from cord blood. The hESC-derived NK cells acquire the ability to lyse human tumor cells by both direct cell-mediated cytotoxicity and Ab-dependent cellular cytotoxicity. Additionally, activated

hESC-derived NK cells up-regulate cytokine production. hESC-derived lymphoid progenitors provide a novel means to charac- Downloaded from terize speciﬁc cellular and molecular mechanisms that lead to development of speciﬁc human lymphocyte populations. These cells may also provide a source for innovative cellular immune therapies. The Journal of Immunology, 2005, 175: 5095–5103.

emopoietic cells can be derived from human embryonic ESC-derived NK cells express CD94/NKG2 receptors in an or- stem cells (hESCs)3 allowed to differentiate either by derly and nonstochastic manner; however, they do not express the stromal cell coculture or formation of embryoid bodies Ly49 receptors, which are analogous to the KIRs found in humans H http://www.jimmunol.org/ (1–4). Analyses of transcription factor and cell surface Ag expres- (15). In contrast, mature NK cells isolated from adult mice express sion suggest that hematopoiesis from hESCs follows developmen- both CD94/NKG2 receptors and Ly49 (16). For human hemopoi- tal kinetics similar to what is observed during normal human on- etic cells derived from more mature sources, acquisition of KIR togeny (1–3, 5). To date, most studies have characterized myeloid, expression in vitro appears to be dependent on the stromal cell line erythroid, and megakaryocytic cells derived from hESCs (1–5). used to support NK differentiation. NK cells cocultured on MS-5 Although the process of lymphopoiesis has been well studied start- stromal cells require IL-21 for KIR expression, whereas NK cells ing from hemopoietic stem cell populations isolated from mouse or cocultured on AFT024 cells do not share this requirement (17, 18). human bone marrow or human umbilical cord blood (UCB) (6–9), In these studies, we report that hESCs can efﬁciently give rise to

considerably less is known about the ability of hESCs to differ- NK cells that express both KIRs and CD94/NKG2a, similar to by guest on September 23, 2021 entiate into the lymphoid lineage. what is observed for mature NK cells found in vivo. More impor- NK cells form a central component in the immune defense tantly, the hESC-derived NK cells exhibit appropriate functional against pathogens and various tumors (10). Putative NK cells and characteristics as displayed by ability to lyse cells by two separate B cells have been identified in cultures of differentiated hESCs (3, mechanisms: direct cell-mediated cytotoxicity and Ab-dependent 11). However, these NK cells were characterized solely on basis of cell-mediated cytoxicity (ADCC). These hESC-derived NK cells CD56 expression, without functional analysis. In addition to CD56 also can be induced to produce cytokines, another hallmark of NK expression, mature NK cells typically express inhibitory and acti- cells. vating receptors, and the balance of signals derived from these receptors regulate NK cell activity. Killer cell Ig-like receptors Materials and Methods (KIRs) and CD94/NKG2 heterodimers are two major classes of Cell culture receptors that interact with MHC class I molecules on target cells as their ligands to specify NK cell activity (12–14). Analysis of The hESC line H9 (obtained from Wicell) was maintained as undifferen- NK cells derived from mouse ESCs has been instructive. Mouse tiated cells as described previously (1, 19). Briefly, undifferentiated hESCs were cocultured with mouse embryonic fibroblasts in DMEM:Ham’s F-12 (Invitrogen Life Technologies) supplemented with 15% knockout serum replacer (Invitrogen Life Technologies), 1% MEM-nonessential amino ac- *Stem Cell Institute and Department of Medicine and †Department of Medicine and Cancer Center, University of Minnesota, Minneapolis, MN 55455 ids (Invitrogen Life Technologies), 1 mM L-glutamine (Mediatech), 0.1 mM 2-ME (Sigma-Aldrich), and 4 ng/ml basic fibroblast growth factor Received for publication May 26, 2005. Accepted for publication August 4, 2005. (Invitrogen Life Technologies). The mouse bone marrow stromal cell line The costs of publication of this article were defrayed in part by the payment of page S17 (kindly provided by Dr. K. Dorshkind, University of California, Los charges. This article must therefore be hereby marked advertisement in accordance Angeles, CA) was grown in DMEM (Invitrogen Life Technologies) con- with 18 U.S.C. Section 1734 solely to indicate this fact. taining 10% FBS, 1% penicillin-streptomycin (P/S) (Invitrogen Life Tech- 1 This work supported in part National Institutes of Health Grant HL-72000 (to nologies), 1% MEM-nonessential amino acids, and 0.1 mM 2-ME. Before D.S.K.) and an American Society of Hematology Scholars Award (to D.S.K.). coculture with hESCs, S17 cells were incubated with conditioned medium ␮ 2 Address correspondence and reprint requests to Dr. Dan S. Kaufman, University of containing 10 g/ml mitomycin C (Bedford Laboratories) before attach- Minnesota, Stem Cell Institute, 420 Delaware Street Southeast, MMC 716, Minne- ment onto gelatin (Sigma-Aldrich)-coated 6-well plates (Nalge Nunc In- apolis, MN 55455. E-mail address: [email protected] ternational). The mouse fetal liver cell line AFT024 (kindly provided by Drs. K. Moore and I. Lemischka, Princeton University, Princeton, NJ) (20) 3 Abbreviations used in this paper: hESC, human embryonic stem cell; ADCC, Ab- dependent cell-mediated cytotoxicity; CFC, colony-forming cell; ESC, embryonic was grown at 33°C in DMEM containing 20% FBS, 1% P/S, and 0.05 mM stem cell; Flt3L, Flt3 ligand; KIR, killer cell Ig-like receptor; NCR, natural cytotox- 2-ME. AFT024 cells were irradiated with 2000 rad before coculture with icity receptor; P/S, penicillin and streptomycin; Q-RT-PCR, quantitative RT-PCR; hESC-derived hemopoietic progenitor cells. UCB was obtained from units SCF, stem cell factor; UCB, umbilical cord blood. that were unacceptable for storage in cord blood banks. The use of all

human tissue was approved by the Committee on the Use of Human Sub- otherwise. ADCC was analyzed by preincubating Raji cells with 4, 1, 0.25, jects in Research at the University of Minnesota. and 0.062 ␮g/ml anti-CD20 Ab (IgG1␬ isotype, rituximab; Genentech) for 30 min. As a negative control, Raji cells preincubated with 4 ␮g/ml IgG1␬ Hemopoietic differentiation of hESCs isotype control Ab (BD Pharmingen) was used. For evaluation of ability to up-regulate IFN-␥ cytokine production, H9 hESCs were cocultured with mouse bone marrow stromal cell line S17, hESC-derived NK cells were incubated in humidified atmosphere at 37°C resulting in H9/S17 cells, as described previously (1, 21). Differentiation and 5% CO with RPMI 1640 medium supplemented with 10% FBS alone medium composed of RPMI 1640 (Mediatech) supplemented with 15% 2 as negative control, 50 ng/ml PMA (Sigma-Aldrich), and 500 ng/ml cal- FBS (HyClone), 2 mM L-glutamine, 0.1 mM 2-ME, 1% MEM-nonessential cium ionophore III (Sigma-Aldrich), as a positive control, or 10 ␮g/ml amino acids, and 1% P/S was changed every 2–3 days. After 14–17 days IL-12 and 100 ␮g/ml IL-18 (R&D Systems). After overnight stimulation, of differentiation, the differentiated hESCs were harvested and made into a cells were incubated with 10 ␮g/ml brefeldin A (Sigma-Aldrich) for 5 h. single-cell suspension using collagenase type IV (Invitrogen Life Technol- Cell surface Ags were first stained for CD56-PE and CD45-allophycocya- ogies), followed by trypsin/EDTA (0.05%; Mediatech) supplemented with nin, in addition to isotype controls, cells were then fixed and permeabilized 2% chick serum (Sigma-Aldrich). Cells were analyzed for hemopoietic (Cytofix/Cytoperm kit; BD Pharmingen), followed by intracellular staining precursor cells by flow cytometry and colony-forming cell (CFC) assay for IFN-␥-FITC (BD Pharmingen). Flow cytometric analysis was per- (1, 21). ϩ formed as described above on the lymphocyte cell population. CD34 Positive selection of CD34ϩ and CD34ϩCD45ϩ cells by UCB-derived NK cells were again used as positive control. magnetic sorting Quantitative real-time PCR analysis of KIR expression Single-cell suspensions from days 14–17 H9/S17 cocultures were prepared as described above. Cell pellet was resuspended in Dulbecco’s PBS (Me- Quantitative real-time PCR (Q-RT-PCR) was preformed as described previously (23, 24). Briefly, total RNA from CD34ϩCD45ϩ H9/S17 and diatech) supplemented with 2% FBS and 1 mM EDTA (Fisher Chemicals) ϩ before magnetic sorting. EasySep CD34 selection kit (StemCell Technol- CD34 UCB cells cultured for 30 days in NK conditions was isolated by Downloaded from ogies) was used to isolate CD34ϩ cells from differentiated hESCs and RNeasy Micro kit (Qiagen), and KIR expression was evaluated using Taq- UCB. For isolation of CD34ϩCD45ϩ cells, the EasySep PE selection kit man probes specific for 13 different KIRs. (StemCell Technologies) was used on CD34ϩ selected cells labeled with CD45-PE (BD Pharmingen). Enrichment was confirmed by flow cytomet- Results ric analysis and typically resulted in 70–90% positive population. Similar Hematopoiesis and NK cell development from hESCs results were obtained by flow cytometric sorting using FACSAria (BD ϩ ϩ ϩ

Biosciences) for CD34 and CD34 CD45 hESC-derived cells. The ability of hESCs to give rise to lymphoid cells was investi- http://www.jimmunol.org/ In vitro generation of NK cells gated using a two-step in vitro differentiation scheme (Fig. 1A). Initially, the hESCs (H9 cell line) were cocultured with the murine Hemopoietic precursor cells were transferred to 24-well plates with a con- bone marrow-derived stromal cell line S17 to derive a heteroge- fluent monolayer of irradiated AFT024 cells in medium designed to max- imize NK cell growth as described previously (22). Briefly, cells were neous population of H9/S17 cells. Consistent with previous find- cocultured in DMEM:Ham’s F-12 supplemented with 20% heat-inacti- ings, these H9/S17 cells contain myeloid progenitor cells (1, 4). vated human AB serum (Nabi), 5 ng/ml sodium selenite (Sigma-Aldrich), After 14–17 days of coculture with S17 cells, myeloid CFCs can 50 ␮M ethanolamine (MP Biomedicals), 20 mg/L ascorbic acid (Sigma- be demonstrated within the differentiated hESC population. Sort- ␮ Aldrich), 25 M 2-ME, 1% P/S, 10 ng/ml IL-15 (PeproTech), 5 ng/ml IL-3 ing for CD34ϩ and CD34ϩCD45ϩ cells results in a significant (PeproTech), 20 ng/ml IL-7 (National Cancer Institute), 20 ng/ml stem cell by guest on September 23, 2021 factor (SCF) (PeproTech), and 10 ng/ml Flt3 ligand (Flt3L) (PeproTech). enrichment in the myeloid CFCs as compared with unsorted H9/ Medium containing fresh cytokines was changed weekly with the excep- S17 cell population (Fig. 1B). tion of IL-3 which was only included for the first week of culture. Wells Both the CD34ϩ and CD34ϩCD45ϩ hESC-derived cells remain were harvested after 7–50 days of NK cell culture, counted for viable cells, heterogeneous, and we hypothesized that these cell populations and assayed for phenotype and function. also contained lymphoid progenitor cells. Two cell populations Flow cytometric analysis identified in UCB and bone marrow as being skewed toward lymphoid differentiation are CD34ϩCD7ϩCD45ϩ and Single-cell suspension of differentiated H9/S17 and hESC-derived NK ϩ ϩ ϩ cells were stained with allophycocyanin, PE-, and FITC-coupled control CD34 CD10 CD45 cells (9). Indeed, these can also be identi- Igs or specific Abs against human blood surface Ags: CD34-APC, CD45- fied in hESCs differentiated on S17 stromal cells (Fig. 1C). To test APC or -PE, CD56-APC or -PE, CD15-PE, CD19-PE, CD33-PE or -FITC, for lymphocyte development, we used a system that has been CD3-FITC, CD158a-FITC, CD158b-FITC, CD158e1-FITC, CD16-FITC, shown previously to support proliferation and differentiation of NKp30-PE, NKp44-PE, NKp46-PE, CD94-FITC, NKG2A-PE (all from ϩ BD Pharmingen), CD158i-PE, and NKG2A-PE (Beckman Coulter). All CD34 cells isolated from bone marrow, peripheral blood, and analyses were performed with a FACSCalibur (BD Biosciences) and UCB cells into NK cells (18, 25). In the present study, sorted FlowJo analysis software (Tree Star). Live cells were identified by pro- CD34ϩ and CD34ϩCD45ϩ cells derived from H9/S17 cells were pidium iodide or 7-aminoactinomycin D exclusion. transferred to a secondary culture with the murine fetal liver-de- NK cell cloning frequency rived AFT024 stromal cell line in medium supplemented with IL- 15, IL-3, IL-7, SCF, and Flt3L (referred to as “NK cell condi- For analysis of frequency of hemopoietic precursor cells with NK cell potential, CD34ϩ and CD34ϩCD45ϩ hESC-derived cells, or CD34ϩ cells tions”). Proliferation of the hESC-derived progenitor cells from the isolated from UCB, were plated in limiting dilutions in 96-well plates with distinct starting cell populations was monitored by harvesting and a confluent monolayer of irradiated AFT024 cells (22). Cells were exposed counting cells cultured in NK cell conditions. Under these condi- to the same NK cell culture conditions as described above. Wells were tions, UCB-derived CD34ϩ cells expanded over 1000-fold, monitored weekly for visual observation of growth. After 30 days of in- whereas CD34ϩCD45ϩ H9/S17 cells demonstrated a significant cubation, NK cell development was assessed from all wells showing visual ϩ ϩ expansion of ϳ40-fold when cultured in these same NK cell con- evidence of growth by flow cytometric analysis for CD56 CD45 cells. ϩ Frequency of NK-potent cells was calculated by Poisson distribution based ditions (Fig. 2A). CD34 H9/S17 cells demonstrated less expan- on number of wells with confirmed growth of NK cells after 30 days of sion, although hemopoietic-like clusters of cells growing in a sim- culture (22). ilar pattern and displaying similar morphology to what was seen ϩ ϩ ϩ Functional evaluation of hESC-derived NK cells for CD34 CD45 hESCs and CD34 UCB cells consistently is

51 found within this cell population (Fig. 2B). These results suggest Direct cytotoxicity assays were performed by standard 4-h Cr release ϩ ϩ assay using the K562 (American Type Culture Collection) and Raji (Amer- that the CD34 CD45 cell population is more enriched in hemo- ican Type Culture Collection) cell lines as target cells (22). Effector cells poietic progenitors responsive to proliferation by cytokines, as ϩ were added in limiting dilution starting at 10:1 E:T ratio unless noted compared with the CD34 cell population derived from hESCs. The Journal of Immunology 5097 Downloaded from

FIGURE 1. In vitro hematopoiesis from hESCs. A, Schema to derive NK cells from hESCs. hESCs are first allowed to differentiate on S17 stromal cells for 14–17 days to derive hemopoietic progenitor cells. These progenitors are then sorted based on CD34 or CD34/45 surface expression and cultured on AFT024 cells with defined cytokines. Development of NK cells was analyzed at specific time points after culture on AFT024 cells. B, Generation of erythroid and myeloid progenitors was analyzed by hemopoietic CFC assay with H9 hESCs allowed to differentiate on S17 stromal cells for 14–17 days. In the present study, colonies produced by unsorted or sorted hESC-derived cell populations were quantified after 14 days (results are mean Ϯ SD of five http://www.jimmunol.org/ separate experiments). C, Flow cytometric analysis of unsorted, CD34ϩ sorted and CD34ϩCD45ϩ sorted H9/S17 cells, in addition to CD34ϩ UCB cells isolated by magnetic sorting.

However, because the proliferation of hESC-derived progenitors is less than that of the UCB cells, it is likely that even this hESC- derived CD34ϩCD45ϩ cell population remains more heterogeneous than CD34ϩ cells isolated from UCB (of note, most CD34ϩ

UCB cells also coexpress CD45). by guest on September 23, 2021

Phenotype and clonal frequency of NK cells derived from hESCs To determine phenotype of the hESC-derived cells cultured in NK cell conditions we analyzed cells after 14, 21, and 28 days by flow cytometry for cell surface expression of CD56, a cell surface marker expressed on human NK cells. Although CD34ϩ H9/S17 cells have a limited expansion when cultured in NK conditions, they demonstrate a robust ability to differentiate into NK cells. After 14 days of culture, Ͼ90% of the cells express CD45, a pan- hemopoietic cell marker, but few CD56ϩ cells are observed (Fig. 3A). By 21 days of culture, a distinct CD56ϩCD45ϩ cell population is observed (14.9%), which increases to 37.5% after 28 days of culture. Similar results are observed for CD34ϩCD45ϩ cells derived from H9/S17 cells (Fig. 3B), suggesting that both CD34ϩ and CD34ϩCD45ϩ cell populations contain hemopoietic progenitors with NK cell potential. As these cultures also led to development of a significant population of CD45ϩCD56Ϫ cells, we also analyzed the cells cultured in NK cell conditions for expression of markers of other hemopoietic lineages (Fig. 4). For CD34ϩ and CD34ϩCD45ϩ H9/S17 cells, no significant expression of CD34 (immature cells), CD19 (B cells), or CD3 (T cells) was observed after 30 days in NK cell conditions. A population of the cells do express CD33 a marker of FIGURE 2. Proliferation of hESC-derived cells cultured in NK condi- ϩ ϩ ϩ immature and mature myeloid cells. However, no detectable ex- tions. A, Proliferation of sorted CD34 (E) and CD34 CD45 (‚) H9/S17 ϩ ϩ ϩ pression of CD15 cells (granulocytes) (Fig. 4) or glycophorinA cells was analyzed after 13, 20, and 30 days in NK cell culture. CD34 UCB cells were used as positive control (Ⅺ) (results are mean Ϯ SD of (CD235a) cells (erythrocytes) (data not shown) was observed after three experiments). B, Images of cell development and proliferation after 30 days of culture. 21 days in NK cell culture (top row: ϫ20 original magnification; bottom To determine the clonal frequency of NK cell progenitors that ϩ ϩ ϩ ϩ row: ϫ100 original magnification). White arrows indicate hemopoietic cell give rise to the CD56 cells, CD34 and CD34 CD45 H9/S17 clusters. cells were plated in limiting dilutions and cultured for 30 days in 5098 NK CELLS DERIVED FROM hESCs

FIGURE 3. Generation of NK cells from hESCs. Flow cytometric analysis of sorted CD34ϩ H9/S17 cells and CD34ϩ cells isolated from UCB (positive control) (A), and CD34ϩCD45ϩ H9/S17 cells (B) cocultured with AFT024 stromal cells in medium supplemented with SCF, Flt3L, IL-3, IL-7, and IL-15 (NK cell conditions) (B) for the indicated number of days. Represen- tative results from three separate experiments are shown. Quadrant markers were set based on controls in Downloaded from which Ͻ0.5% of the cells were included in the quadrants. Exclusion of propidium iodide was used for gating on live cells. http://www.jimmunol.org/ by guest on September 23, 2021

NK cell conditions. After 30 days, wells with visual growth were that express only low levels of CD158e1 and do not express harvested and analyzed by flow cytometry for presence of CD56ϩ CD158i. The KIR protein expression as analyzed by flow cytom- cells. We found CD34ϩ H9/S17 cells to have a NK cell frequency etry was further resolved by a Q-RT-PCR method to better define of 0.16% (Table I). However, sorting for CD34ϩCD45ϩ H9/S17 the expression of 13 individual KIR genes (23, 24). This Q-RT- cells significantly increased the NK cell cloning frequency (1.1%) PCR analysis showed hESC-derived NK cells express transcripts to a level comparable to the frequency observed for CD34ϩ UCB for KIR2DS1, KIR2DL4, KIR2DL5, KIR2DS5, KIR3DS1, and cells cultured in the same manner (2.4%). KIR3DL2, in addition to KIRs mentioned above, as determined by flow cytometry (our unpublished observations). Furthermore, be- hESC-derived NK cells express inhibitory and activating cause the Ab used to detect CD158b does not specifically distin- receptors guish between KIR2DL2, KIR2DS2, and KIR2DL3, the Q-RT- As initial phenotypic analysis suggested that hESCs can differen- PCR assay resolved that the hESC-derived NK cells expressed tiate into NK cells, we next characterized the hESC-derived cells only KIR2DL3 to account for the CD158b surface expression. for surface expression of other NK cell Ags. NK cell cytolytic The hESC-derived NK cells also express the C-type lectin-like activity is regulated by signals initiated by specific activating and receptors CD94 and NKG2A (Fig. 5C), which upon dimerization inhibitory receptors (16). One important family of receptors in- form an inhibitory receptor that primarily interacts with HLA-E volved in the regulation of cytolytic activity is the KIRs. Initially, (26). Another group of activating receptors, collectively termed we analyzed the expression of four KIRs using a mixture of Abs natural cytotoxicity receptors (NCRs), can be identified by their specific for three inhibitory KIRs: KIR2DL1 (CD158a), exclusive expression on NK cells (27–29). These include NKp30, KIR2DL2/DL3/DS2 (CD158b), and KIR3DL1 (CD158e1) and NKp44, and NKp46, all of which are expressed on the hESC- one activating KIR, KIR2DS4 (CD158i). During NK culture of derived NK cells (Fig. 5C). In addition, we found hESC-derived CD34ϩCD45ϩ H9/S17 cells, CD56ϩ NK cells start to express NK cells express lymphoid associated markers CD7 and CD2 (Fig. KIRs after 18 days of culture (Fig. 5A). After 50 days, 40% of cells 6). In vivo, most mature NK cells express CD16. However, in vitro are CD56ϩKIRϩ cells, suggesting a time-dependent up-regulation studies of UCB-derived NK cell differentiation by coculture with of KIR expression (Fig. 5A). Interestingly, the KIR expression is the MS-5 stromal cell line depends on the addition of IL-21 to consistently higher in the hESC-derived NK cells as compared differentiate to fully mature CD56ϩCD16ϩ NK cells (17). Inter- with NK cells derived from CD34ϩ UCB cells. When investigat- estingly, this is not a requirement of these studies using the ing the expression of the individual KIRs, hESC-derived NK cells AFT024 stromal cell line because hESC- and UCB-derived NK express CD158b, CD158e1, and CD158i but do not express cells express CD16 after 30 days of culture (Fig. 5C). Notably, a CD158a (Fig. 5B). This is different from UCB-derived NK cells greater percentage of the hESC-derived CD56ϩ cells coexpressed The Journal of Immunology 5099

Table I. NK cell clonal frequency

CD34ϩ hESC CD34ϩCD45ϩ hESC CD34ϩ UCB

Frequency 1:615 1:93 1:42 95% conﬁdence (433;873) (66;132) (29;59) interval

CD34ϩ or CD34ϩCD45ϩ cells derived from hESCs allowed to differentiate on S17 cells or CD34ϩ cells isolated from UCB were cultured in limiting dilution in NK cell conditions for 30 days. Wells were then evaluated for cell proliferation, and presence of CD56ϩCD45ϩ NK cells was conﬁrmed by ﬂow cytometry. Poisson distribution was used to calculate frequency of NK progenitor cells (22).

lytic activity at this time (Fig. 7A) due to the few CD56ϩCD45ϩ NK cells present at this early time point (Fig. 3A). However, at day 32, the CD56ϩCD45ϩ cells have developed into a signiﬁcant population, and these cells demonstrate cytolytic activity similar to UCB-derived NK cells. CD34ϩCD45ϩ H9/S17-derived NK cells

also demonstrate cytolytic activity after 30 days of NK culture Downloaded from comparable to the cytolytic activity observed for UCB-derived NK cells (Fig. 7B). The cytolytic activity observed for the hESC-derived cells reside in the CD56ϩ population, as sorting for these cells after 30 days of culture markedly enhances the cytolytic activity compared with the unsorted cell population (Fig. 7C). Fur-

thermore, no signiﬁcant cytolytic activity was observed in the http://www.jimmunol.org/ CD56Ϫ cell population, even at higher E:T ratios. Similar results were observed for the UCB-derived NK cells (data not shown). As the hESC-derived NK cells express CD16 (Fc␥RIII), a receptor that binds the Fc region of IgG molecules (Fig. 5C), we also tested their ability to mediate ADCC. The cytolytic activity of hESC-derived NK cells was targeted against the NK-resistant Raji cell line incubated with anti-CD20 Ab or isotype control Ab. Our results show that the hESC-derived NK cells can mediate lysis of

Raji cells in an anti-CD20 dose-dependent manner, whereas Raji by guest on September 23, 2021 cells incubated without Ab or isotype control Ab were not effectively lysed (Fig. 7D).

Cytokine production from hESC-derived NK cells Another means to analyze functional characteristics of NK cells is their ability to up-regulate IFN-␥ production in response to IL-12/ FIGURE 4. Cell surface expression of myeloid and lymphoid Ags. IL-18 stimulation. Consistent with this capacity, hESC-derived Sorted CD34ϩ and CD34ϩCD45ϩ H9/S17 cells were cultured in NK cell NK cells stimulated overnight with IL-12 and IL-18 up-regulated conditions for 30 days and analyzed by flow cytometry for cell surface ␥ expression of markers for hemopoietic precursor/progenitors (CD34), gran- IFN- production in the same manner as when these cells are stim- ulocytes (CD15), B lymphocytes (CD19), T lymphocytes (CD3), and my- ulated by PMA and calcium ionophore, similar to what seen in eloid cells (CD33). CD34ϩ cells isolated from UCB were used as a positive UCB-derived NK cells used as positive controls (Fig. 8). control. Quadrant markers were set based on controls in which Ͻ0.5% of the cells were included in the quadrants. Exclusion of propidium iodide Discussion was used for gating on live cells. In the present study, we demonstrate that hESCs can develop into mature, functional NK cells. These hESC-derived NK cells acquire KIR expression similar to what is observed for mature NK cells in ϩ CD16, as compared with the CD56 cells derived from UCB vivo (16). CD94/NKG2A, NCRs, and CD16 are also expressed by progenitors. the hESC-derived NK cells. Sorting for specific cell surface phe- notypes to isolate hESC-derived hemopoietic progenitors results in Cytolytic activity of hESC-derived NK cells a significant increase in NK cell cloning frequency. More impor- The above phenotypic results suggest that hESCs can be efficiently tantly, hESC-derived NK cells exhibited cytokine production and induced to differentiate into NK cells in vitro. We next examined cytolytic activity against human tumor cells by both direct cell- functional cytolytic activity of the hESC-derived NK cells. One mediated cellular cytotoxicity and ADCC. These results suggest hallmark of NK cells is their ability to target and lyse human tumor that not only can functional lymphoid cells be routinely and effi- cells (30). Thus, hESC-derived NK cells were harvested and tested ciently derived from undifferentiated hESCs but also specific phe- for their cytolytic activity toward K562 erythroleukemia cells in a notypic cell populations derived from hESCs can be effectively standard 4-h 51Cr release assay. Cells derived from UCB after 17 and efficiently selected, expanded, and induced to mature into spe- days of NK cell culture on AFT024 cells supplemented with de- cific blood lineages. fined cytokines are able to effectively kill K562 cells. As expected, Although recent reports suggest that phenotypic NK cells can be CD34ϩ H9/S17-derived cells did not display any significant cyto- derived from hESCs, prior analyses have relied solely on CD56 5100 NK CELLS DERIVED FROM hESCs

FIGURE 5. hESC-derived NK cells express KIRs, CD94/NKG2a, and NCRs. A, Flow cytometric analysis of KIR acquisition (pooled Abs against CD158a, CD158b, CD158e1, and CD158i) in sorted CD34ϩCD45ϩ H9/S17 cells cultured in NK cell conditions for the indicated number of days. B and C, Sorted CD34ϩCD45ϩ H9/S17 cultured in NK conditions for 30 days were analyzed by ﬂow cytometry for expression of individual KIRs and the C-type lectin-like receptors CD94 and NKG2A, in addition to expression of CD16 (Fc␥RIIIA) and NCRs. Representa- tive results from two separate experiments are shown. CD34ϩ UCB cell were used as a positive control. Downloaded from Quadrant markers were set based on controls in which Ͻ0.5% of the cells were included in the quadrants. Ex- clusion of propidium iodide was used for gating on live cells. Similar results were found using sorted CD34ϩ http://www.jimmunol.org/ H9/S17 cells.

expression as a marker of NK cells (3, 11). CD56 alone is insuf- CD16 expression to generate functional cytolytic and cytokine- ficient to identify NK cells because it is promiscuously expressed producing NK cells. CD56 and KIR expression was acquired in a on neuronal (31) and pancreatic cells (32), as well as in low levels time-dependent manner, that agrees with current models of NK on undifferentiated hESCs (P. S. Woll and D. S. Kaufman, unpub- cell maturation (16). by guest on September 23, 2021 lished observations). CD56 is also found on myeloid cells in some The molecular and cellular events that regulate development of patients with chronic myeloid leukemia (33). Thus, to more clearly NK cell precursors to mature NK cells are not well characterized. demonstrate generation of NK cells, characterization of additional However, a sequential pattern of cell surface Ag expression has receptors and NK cell activity is needed. In the present study, we been identified. The earliest cells committed to the NK cell lineage demonstrate hESC-derived NK cells acquire CD94, KIR, and can be identified as CD161ϩCD56Ϫ, which subsequently give rise to CD56ϩ NK cells. CD94 expression is acquired before KIR and CD16 expression, generating cytolytic and cytokine-producing NK cells (16). Furthermore, NK cells can be classified either by CD56 expression pattern or cytokine production. CD56bright cells have low expression of CD16 and KIRs, poor cytolytic activity, and high levels of cytokine production (34, 35). In contrast, most cytolytic activity is found in the CD56dim cells, which also have high surface expression of CD16 and KIRs. Recent reports suggest that CD56bright cells are less mature than CD56dim cells (36, 37). In addition, cytokine production has been used to classify NK cell populations. A linear developmental progression from IL- 13ϩIFN-␥Ϫ stage cells (type 2) to an intermediate IL-13ϩIFN-␥ϩ stage (type 0), followed by IL-13ϪIFN-␥ϩ cells (type 1), has been suggested (38). The development of hESC-derived NK cells closely recapitu- lates normal NK cell developmental kinetics. This correlation strongly suggests that the hESC system provides an accurate developmental model to evaluate specific cellular and genetic mech- FIGURE 6. hESC-derived NK cells express the lymphoid associated ϩ ϩ ϩ anisms that regulate NK cell maturation. One unique aspect of the Ags CD2 and CD7. Sorted CD34 and CD34 CD45 H9/S17 cells cul- ESC system is that the differentiation process follows distinct se- tured in NK cell conditions for 30 days were analyzed by flow cytometry quential steps of hemopoietic maturation that can be monitored at for cell surface expression of the lymphoid-associated Ags CD7 and CD2. ϩ very early developmental stages. Initial differentiation of hESCs CD34 cells isolated from UCB were used as positive control. Quadrant ϩ markers were set based on controls in which Ͻ0.5% of the cells were can generate CD34 cells that give rise to myeloid and lymphoid included in the quadrants. Exclusion of propidium iodide was used for progenitors, which in turn produce mature blood cells. Unlike gating on live cells. UCB and bone marrow, hESCs do not initially contain mature The Journal of Immunology 5101 Downloaded from

FIGURE 7. Functional NK cells derived from hESCs. A, Cytolytic activity in sorted CD34ϩ H9/S17-derived cells (F) or CD34ϩ UCB-derived cells (f) cultured in NK cell conditions for 17 and 32 days were analyzed for ability to lyse K562 cells by standard 4-h 51Cr release assay. B, CD34ϩCD45ϩ H9/S17-derived cells (Œ) or CD34ϩ UCB-derived cells (f) cultured in NK conditions for 30 days evaluated for ability to lyse K562 cells (results are mean Ϯ SD of three separate experiments). C, CD56ϩ (ϫ) and CD56Ϫ (ࡗ) cells derived from CD34ϩCD45ϩ H9/S17 cells after 35 days in NK cell conditions were isolated and tested for ability to lyse K562 cells and compared with the unsorted cell population (Œ). Similar results were found for CD34ϩ UCB cells cultured in same conditions. D, ADCC was tested by preincubating Raji cells with indicated concentrations of anti-CD20 Ab (␣CD20) or with 4 ␮g/ml http://www.jimmunol.org/ IgG1k isotype control Ab. Cells derived from CD34ϩCD45ϩ H9/S17 and CD34ϩ UCB cultured in NK conditions for 30 days were used as effectors against the Ab-treated Raji cells at a 20:1 E:T ratio. hemopoietic cells, reducing the possibility that contaminating ma- Interestingly, the KIR expression on the hESC-derived NK cells ture cells might obscure in vitro analysis of differentiation path- was higher than on UCB-derived NK cells. Because of the highly ways. Thus, following the multistep hemopoietic differentiation polymorphic nature of the KIR locus on chromosome 19, this could process from hESCs allows for an unbiased and reproducible anal- be explained by differences in inherited KIR genes (23). However, ysis of transcriptional regulation of differentiation and maturation. genetic heterogeneity is unlikely to be the sole explanation for this by guest on September 23, 2021 As far as the models of NK cell maturation based on levels of phenomenon, as the same difference in KIR expression was ob- CD56 expression and cytokine production (38, 39), our results do served when NK cells derived from multiple UCB-donors were not yet provide enough information to evaluate if one pathway compared with that of hESC-derived NK cells. Instead, our results predominates for hESC-derived NK cells. However, the cytolytic support previous findings where KIR acquisition on developing ␥ activity and IFN- cytokine production demonstrated in the hESC- NK cells in vitro inversely correlated with the ontogeny of the derived NK cells suggest that the hESCs can now serve as a model stem cell source (18). This has been explained previously by the system to distinguish these two models of NK cell maturation. relatively higher proliferation capabilities of the more immature source. However, this is not a likely explanation to describe our results, as the proliferation observed for hESCs-derived NK cells was lower compared with the proliferation observed for UCB cells (Fig. 2A). This may be due to the hESC-derived CD34ϩCD45ϩ cell population remaining more heterogeneous and containing quantitatively fewer lymphocyte progenitors than CD34ϩ cells isolated from UCB. Certainly, this heterogeneity between hESC- and UCB-derived progenitors needs to be compared because these studies so far have been incomplete. Although phenotypic analysis of hESCs differentiated on OP9 stromal cells suggests that CD34ϩ hESC-derived cells resemble primitive bone marrow and intraembryonic hemopoietic precursors by expression of CD90, CD117, and CD164, the functional relevance of this remains unknown (3). Indeed, we can identify CD34ϩCD45ϩCD7ϩ and CD34ϩCD45ϩCD10ϩ FIGURE 8. hESC-derived NK cells up-regulate IFN-␥ cytokine pro- cells from differentiated hESCs, corresponding to more mature duction following stimulation with IL-12 and IL-18. Sorted CD34ϩCD45ϩ common lymphoid progenitor cell populations identified in UCB H9/S17-derived cells cultured in NK conditions for 30 days were stimu- and bone marrow (9) (Fig. 1C). Future studies will determine the lated overnight with medium alone, PMA, and calcium ionophore III NK cell cloning frequency and proliferative potential of these (PMAϩCa/I) or IL-12 and IL-18. Cells were analyzed by intracellular flow hESC-derived cell populations, and comparison to similar popu- cytometric straining for IFN-␥. Quadrant markers were set based on controls in which Ͻ0.5% of the cells were included in the quadrants. Repre- lations isolated from UCB and bone marrow will be instructive for sentative intracellular flow cytometric analysis for three separate experi- establishing a hemopoietic maturation scheme from hESC-derived ments is shown. hemopoietic progenitors. 5102 NK CELLS DERIVED FROM hESCs

As MHC class I molecules act as ligands for KIRs, some studies References suggest that MHC class I exposure might affect NK cell develop- 1. Kaufman, D. S., E. T. Hanson, R. L. Lewis, R. Auerbach, and J. A. Thomson. ment and KIR expression (40). Mice lacking MHC class I have a 2001. Hematopoietic colony-forming cells derived from human embryonic stem cells. Proc. Natl. Acad. Sci. USA 98: 10716–10721. higher expression of Ly49 receptors than wild-type littermates 2. Chadwick, K., L. Wang, L. Li, P. Menendez, B. Murdoch, A. Rouleau, and (41). In humans, although HLA and KIR genes are not linked, M. Bhatia. 2003. Cytokines and BMP-4 promote hematopoietic differentiation of HLA class I appears to affect the frequency of KIR expression on human embryonic stem cells. Blood 102: 906–915. 3. Vodyanik, M. A., J. A. Bork, J. A. Thomson, and I. I. Slukvin. 2005. Human developing NK cells that reconstitute after hemopoietic cell trans- embryonic stem cell-derived CD34ϩ cells: efficient production in the coculture plantation (40). Another possible explanation to account for varied with OP9 stromal cells and analysis of lymphohematopoietic potential. Blood 105: 617–626. level of expression of NK cell surface receptors between different 4. Tian, X., J. K. Morris, J. L. Linehan, and D. S. Kaufman. 2004. Cytokine re- cell populations is the epigenetic regulation of KIRs. Recently, the quirements differ for stroma and embryoid body-mediated hematopoiesis from expression of KIRs has been found to be regulated by the meth- human embryonic stem cells. Exp. Hematol. 32: 1000–1009. 5. Zambidis, E. T., B. Peault, T. S. Park, F. Bunz, and C. I. Civin. 2005. Hemato- ylation status of the KIR locus (42, 43). As hESCs are associated poietic differentiation of human embryonic stem cells progresses through sequen- with a more open chromatin structure (44), it is possible that this tial hemato-endothelial, primitive, and definitive stages resembling human yolk epigenetic regulation is less efficient in the hESC-derived NK sac development. Blood 106: 860–870. 6. Kondo, M., I. L. Weissman, and K. Akashi. 1997. Identification of clonogenic cells. Future analyses will be important to further investigate this common lymphoid progenitors in mouse bone marrow. Cell 91: 661–672. issue. 7. Galy, A., M. Travis, D. Cen, and B. Chen. 1995. Human T, B, natural killer, and dendritic cells arise from a common bone marrow progenitor cell subset. Immu- During embryonic development, hematopoiesis occurs in two nity 3: 459–473. waves. The primitive hemopoietic cells are confined to the ex- 8. Hao, Q. L., J. Zhu, M. A. Price, K. J. Payne, L. W. Barsky, and G. M. Crooks. traembryonic yolk sac, generating primarily nucleated RBCs. 2001. Identification of a novel, human multilymphoid progenitor in cord blood. Downloaded from Blood 97: 3683–3690. These cells lack in vitro lymphomyeloid cell potential, which is 9. Haddad, R., P. Guardiola, B. Izac, C. Thibault, J. Radich, A. L. Delezoide, found in later definitive hemopoietic cells (45). Demonstration of C. Baillou, F. M. Lemoine, J. C. Gluckman, F. Pflumio, and B. Canque. 2004. lymphocytes from hESCs would indicate that these cells are ca- Molecular characterization of early human T/NK and B-lymphoid progenitor cells in umbilical cord blood. Blood 104: 3918–3926. pable of definitive hematopoiesis. However, as phenotypic NK 10. Pelayo, R., R. Welner, S. S. Perry, J. Huang, Y. Baba, T. Yokota, and cells can be derived from yolk sac (46), the results presented here P. W. Kincade. 2005. Lymphoid progenitors and primary routes to becoming cells of the immune system. Curr. Opin. Immunol. 17: 100–107. cannot be solely used to distinguish primitive vs definitive hema- http://www.jimmunol.org/ ϩ 11. Zhan, X., G. Dravid, Z. Ye, H. Hammond, M. Shamblott, J. Gearhart, and topoiesis from the hESCs. The demonstration of CD19 B cells L. Cheng. 2004. Functional antigen-presenting leucocytes derived from human derived from hESCs, a population not found to be derived from embryonic stem cells in vitro. Lancet 364: 163–171. 12. Colonna, M., and J. Samaridis. 1995. Cloning of immunoglobulin-superfamily yolk-sac progenitors (3, 46), and other recent studies that demon- members associated with HLA-C and HLA-B recognition by human natural killer strate expression of ␥ and ␤ globin genes in hESC-derived eryth- cells. Science 268: 405–408. rocytes (5, 47) together more conclusively establishes definitive 13. D’Andrea, A., C. Chang, K. Franz-Bacon, T. McClanahan, J. H. Phillips, and L. L. Lanier. 1995. Molecular cloning of NKB1: a natural killer cell receptor for hemopoietic cells can be derived from hESCs. Indeed, recent stud- HLA-B allotypes. J. Immunol. 155: 2306–2310. ies also more clearly delineate that a transition from hemopoietic- 14. Lee, N., M. Llano, M. Carretero, A. Ishitani, F. Navarro, M. Lopez-Botet, and D. E. Geraghty. 1998. HLA-E is a major ligand for the natural killer inhibitory endothelial cell precursors to primitive then definitive hemopoietic receptor CD94/NKG2A. Proc. Natl. Acad. Sci. USA 95: 5199–5204. by guest on September 23, 2021 cells can be modeled with hESCs (5, 48). 15. Lian, R. H., M. Maeda, S. Lohwasser, M. Delcommenne, T. Nakano, In addition to extensive in vitro characterization of the hemo- R. E. Vance, D. H. Raulet, and F. Takei. 2002. Orderly and nonstochastic acquisition of CD94/NKG2 receptors by developing NK cells derived from em- poietic potential of hESCs, recent reports demonstrate in vivo en- bryonic stem cells in vitro. J. Immunol. 168: 4980–4987. graftment of hemopoietic cells derived from hESCs when trans- 16. Colucci, F., M. A. Caligiuri, and J. P. Di Santo. 2003. What does it take to make planted into immunodeficient mice or fetal sheep (49–51). Similar a natural killer? Nat. Rev. Immunol. 3: 413–425. 17. Sivori, S., C. Cantoni, S. Parolini, E. Marcenaro, R. Conte, L. Moretta, and results have also been obtained from transplantation of hemopoi- A. Moretta. 2003. IL-21 induces both rapid maturation of human CD34ϩ cell etic cells derived from cynomolgus monkey ESCs transplanted precursors towards NK cells and acquisition of surface killer Ig-like receptors. Eur. J. Immunol. 33: 3439–3447. into fetal sheep (52). These studies of in vivo engraftment remain 18. Miller, J. S., and V. McCullar. 2001. Human natural killer cells with polyclonal the gold standard to evaluate the full hemopoietic potential for lectin and immunoglobulin-like receptors develop from single hematopoietic specific precursor cell populations. stem cells with preferential expression of NKG2A and KIR2DL2/L3/S2. Blood 98: 705–713. Future clinical application of hESCs may involve use of these 19. Thomson, J. A., J. Itskovitz-Eldor, S. S. Shapiro, M. A. Waknitz, J. J. Swiergiel, cells an alternative source of hemopoietic cells of various specific V. S. Marshall, and J. M. Jones. 1998. Embryonic stem cell lines derived from lineages, including hemopoietic stem cells, mature erythroid cells, human blastocysts. Science 282: 1145–1147. 20. Moore, K. A., H. Ema, and I. R. Lemischka. 1997. In vitro maintenance of highly platelets, and lymphocytes (53–55). NK cells provide important purified, transplantable hematopoietic stem cells. Blood 89: 4337–4347. cell-mediated antitumor activity, as clearly demonstrated in recent 21. Tian, X., and D. S. Kaufman. 2004. Hematopoietic development of human embryonic stem cells in culture. Methods Mol. Med. 105: 425–436. clinical trials (56, 57). Because hESC-derived NK cells demon- 22. Miller, J. S., V. McCullar, M. Punzel, I. R. Lemischka, and K. A. Moore. 1999. strate mature effector functions, these cells may prove to be useful Single adult human CD34ϩ/LinϪ/CD38Ϫ progenitors give rise to natural killer in clinical therapeutic applications that require further cells, B-lineage cells, dendritic cells, and myeloid cells. Blood 93: 96–106. 23. Uhrberg, M., N. M. Valiante, B. P. Shum, H. G. Shilling, K. Lienert-Weidenbach, investigation. B. Corliss, D. Tyan, L. L. Lanier, and P. Parham. 1997. Human diversity in killer cell inhibitory receptor genes. Immunity 7: 753–763. 24. Xiao, F., M. Pitt, R. Albrecht, V. McCullar, K. Brungaard, and J. S. Miller. 2003. A novel RT-PCR assay to measure comparative KIR expression at the single gene Acknowledgments level in donor recipient pairs of unrelated transplants. Blood 102: 723 (Abstr.). We gratefully acknowledge Jon Linehan and Julie Morris for help with 25. Miller, J. S., C. Verfaillie, and P. McGlave. 1992. The generation of human natural killer cells from CD34ϩ/DRϪ primitive progenitors in long-term bone maintaining cells lines, Sarah McNearney and Valarie McCullar for tech- marrow culture. Blood 80: 2182–2187. nical support, Dr. Paul Leibson for expert advice, Paul Marker for 26. Braud, V. M., D. S. Allan, C. A. O’Callaghan, K. Soderstrom, A. D’Andrea, FACSAria sorting, Dr. Catherine Verfaille, and lab members for sharing G. S. Ogg, S. Lazetic, N. T. Young, J. I. Bell, J. H. Phillips, L. L. Lanier, and reagents and equipment. A. J. McMichael. 1998. HLA-E binds to natural killer cell receptors CD94/ NKG2A, B, and C. Nature 391: 795–799. 27. Pende, D., S. Parolini, A. Pessino, S. Sivori, R. Augugliaro, L. Morelli, E. Marcenaro, L. Accame, A. Malaspina, R. Biassoni, et al. 1999. Identification and molecular characterization of NKp30, a novel triggering receptor involved in Disclosures natural cytotoxicity mediated by human natural killer cells. J. Exp. Med. 190: The authors have no financial conflict of interest. 1505–1516. The Journal of Immunology 5103

28. Vitale, M., C. Bottino, S. Sivori, L. Sanseverino, R. Castriconi, E. Marcenaro, maintains allele-specific KIR gene expression in human natural killer cells. R. Augugliaro, L. Moretta, and A. Moretta. 1998. NKp44, a novel triggering J. Exp. Med. 197: 245–255. surface molecule specifically expressed by activated natural killer cells, is in- 43. Santourlidis, S., H. I. Trompeter, S. Weinhold, B. Eisermann, K. L. Meyer, volved in non-major histocompatibility complex-restricted tumor cell lysis. P. Wernet, and M. Uhrberg. 2002. Crucial role of DNA methylation in determi- J. Exp. Med. 187: 2065–2072. nation of clonally distributed killer cell Ig-like receptor expression patterns in NK 29. Sivori, S., M. Vitale, L. Morelli, L. Sanseverino, R. Augugliaro, C. Bottino, cells. J. Immunol. 169: 4253–4261. L. Moretta, and A. Moretta. 1997. p46, a novel natural killer cell-specific surface 44. Rasmussen, T. P. 2003. Embryonic stem cell differentiation: a chromatin per- molecule that mediates cell activation. J. Exp. Med. 186: 1129–1136. spective. Reprod. Biol. Endocrinol. 1: 100. 30. Smyth, M. J., E. Cretney, J. M. Kelly, J. A. Westwood, S. E. Street, H. Yagita, 45. Lensch, M. W., and G. Q. Daley. 2004. Origins of mammalian hematopoiesis: in K. Takeda, S. L. van Dommelen, M. A. Degli-Esposti, and Y. Hayakawa. 2005. vivo paradigms and in vitro models. Curr. Top. Dev. Biol. 60: 127–196. Activation of NK cell cytotoxicity. Mol. Immunol. 42: 501–510. 46. Tavian, M., C. Robin, L. Coulombel, and B. Peault. 2001. The human embryo, 31. Piper, D. R., T. Mujtaba, M. S. Rao, and M. T. Lucero. 2000. Immunocytochem- but not its yolk sac, generates lympho-myeloid stem cells: mapping multipotent ical and physiological characterization of a population of cultured human neural hematopoietic cell fate in intraembryonic mesoderm. Immunity 15: 487–495. precursors. J. Neurophysiol. 84: 534–548. 47. Cerdan, C., A. Rouleau, and M. Bhatia. 2004. VEGF-A165 augments erythro- 32. Lackie, P. M., C. Zuber, and J. Roth. 1994. Polysialic acid of the neural cell poietic development from human embryonic stem cells. Blood 103: 2504–2512. adhesion molecule (N-CAM) is widely expressed during organogenesis in me- 48. Wang, L., L. Li, F. Shojaei, K. Levac, C. Cerdan, P. Menendez, T. Martin, sodermal and endodermal derivatives. Differentiation 57: 119–131. A. Rouleau, and M. Bhatia. 2004. Endothelial and hematopoietic cell fate of 33. Lanza, F., S. Bi, G. Castoldi, and J. M. Goldman. 1993. Abnormal expression of human embryonic stem cells originates from primitive endothelium with heman- N-CAM (CD56) adhesion molecule on myeloid and progenitor cells from chronic gioblastic properties. Immunity 21: 31–41. myeloid leukemia. Leukemia 7: 1570–1575. 49. Wang, L., P. Menendez, F. Shojaei, L. Li, F. Mazurier, J. E. Dick, C. Cerdan, 34. Lanier, L. L., A. M. Le, C. I. Civin, M. R. Loken, and J. H. Phillips. 1986. The K. Levac, and M. Bhatia. 2005. Generation of hematopoietic repopulating cells relationship of CD16 (Leu-11) and Leu-19 (NKH-1) antigen expression on hu- from human embryonic stem cells independent of ectopic HOXB4 expression. J. Exp. Med. 201: 1603–1614. man peripheral blood NK cells and cytotoxic T lymphocytes. J. Immunol. 136: ϩ 4480–4486. 50. Kaufman, D. S., P. S. Woll, C. H. Martin, J. Linehan, and X. Tian. 2004. CD34 cells derived from human embryonic stem cells demonstrate hematopoietic stem

35. Andre, P., O. Spertini, S. Guia, P. Rihet, F. Dignat-George, H. Brailly, J. Sampol, Downloaded from cell potential in vitro and in vivo. Blood 104(Suppl. 1): 163 (Abstr.). P. J. Anderson, and E. Vivier. 2000. Modiﬁcation of P-selectin glycoprotein 51. Narayan, A. D., J. L. Chase, A. Ersek, J. A. Thomson, R. L. Lewis, D. S. Kauf- ligand-1 with a natural killer cell-restricted sulfated lactosamine creates an alter- man, and E. D. Zanjani. 2004. Human embryonic stem cell-derived hematopoi- nate ligand for L-selectin. Proc. Natl. Acad. Sci. USA 97: 3400–3405. etic elements are capable of engrafting primary as well as secondary fetal sheep 36. Ferlazzo, G., D. Thomas, S. L. Lin, K. Goodman, B. Morandi, W. A. Muller, recipients. Blood 104: 733 (Abstr.). A. Moretta, and C. Munz. 2004. The abundant NK cells in human secondary 52. Sasaki, K., Y. Nagao, Y. Kitano, H. Hasegawa, H. Shibata, M. Takatoku, lymphoid tissues require activation to express killer cell Ig-like receptors and S. Hayashi, K. Ozawa, and Y. Hanazono. 2005. Hematopoietic microchimerism J. Immunol. become cytolytic. 172: 1455–1462. in sheep after in utero transplantation of cultured cynomolgus embryonic stem 37. Jacobs, R., M. Stoll, G. Stratmann, R. Leo, H. Link, and R. E. Schmidt. 1992. Ϫ ϩ cells. Transplantation 79: 32–37. http://www.jimmunol.org/ CD16 CD56 natural killer cells after bone marrow transplantation. Blood 79: 53. Kaufman, D. S., and J. A. Thomson. 2002. Human ES cells: haematopoiesis and 3239–3244. transplantation strategies. J. Anat. 200: 243–248. 38. Loza, M. J., and B. Perussia. 2001. Final steps of natural killer cell maturation: 54. Odorico, J. A., D. S. Kaufman, and J. A. Thomson. 2001. Multilineage differ- a model for type 1-type 2 differentiation? Nat. Immunol. 2: 917–924. entiation from human embryonic stem cell lines. Stem Cells 19: 193–204. 39. Cooper, M. A., T. A. Fehniger, and M. A. Caligiuri. 2001. The biology of human 55. Rosenthal, N. 2003. Prometheus’s vulture and the stem-cell promise. N. Engl. natural killer-cell subsets. Trends Immunol. 22: 633–640. J. Med. 349: 267–274. 40. Shilling, H. G., N. Young, L. A. Guethlein, N. W. Cheng, C. M. Gardiner, 56. Miller, J. S., Y. Soignier, A. Panoskaltsis-Mortari, S. A. McNearney, G. H. Yun, D. Tyan, and P. Parham. 2002. Genetic control of human NK cell repertoire. S. K. Fautsch, D. McKenna, C. Le, T. E. Defor, L. J. Burns, et al. 2005. Suc- J. Immunol. 169: 239–247. cessful adoptive transfer and in vivo expansion of human haploidentical NK cells 41. Held, W., J. R. Dorfman, M. F. Wu, and D. H. Raulet. 1996. Major histocom- in patients with cancer. Blood 105: 3051–3057. patibility complex class I-dependent skewing of the natural killer cell Ly49 re- 57. Ruggeri, L., M. Capanni, E. Urbani, K. Perruccio, W. D. Shlomchik, A. Tosti,

ceptor repertoire. Eur. J. Immunol. 26: 2286–2292. S. Posati, D. Rogaia, F. Frassoni, F. Aversa, M. F. Martelli, and A. Velardi. 2002. by guest on September 23, 2021 42. Chan, H. W., Z. B. Kurago, C. A. Stewart, M. J. Wilson, M. P. Martin, Effectiveness of donor natural killer cell alloreactivity in mismatched hemato- B. E. Mace, M. Carrington, J. Trowsdale, and C. T. Lutz. 2003. DNA methylation poietic transplants. Science 295: 2097–2100.