Negative Selection by Apoptosis Enriches Progenitors in Naïve and Expanded Human Umbilical Cord Blood Grafts

ORIGINAL ARTICLE Negative selection by apoptosis enriches progenitors in naïve and expanded human umbilical cord blood grafts

K Mizrahi1, S Ash2, T Peled3, I Yaniv4, J Stein2 and N Askenasy1

The influence of TNF-α and Fas-ligand (FasL) on viability and function was evaluated in fresh- and expanded-umbilical cord blood (UCB) cells. CD34+ progenitors and T cells display outstanding survival, whereas ~ 30% and >50% B lymphocytes and myeloid cells undergo spontaneous apoptosis within 24 and 48 h, respectively. Although the impact of exposure to toxic doses of FasL and TNF-α was undetectable in measurements of apoptosis; removal of dead cells after 2 days of incubation with the ligands revealed a twofold increase in frequency of colony-forming cells (CFU). The sensitivity of progenitors to apoptosis was also unaffected by Fas cross-linking following TNF-induced upregulation of the receptor, increasing CFU frequency without impairing SCID repopulating cell (SRC) activity. Most significant enrichment in CD34+ progenitors and corresponding increase in CFU frequency were observed when FasL was applied during the final week of ex vivo expansion under the influence of nicotinamide, without impairing SRC activity. These data emphasize differential sensitivities of UCB progenitors and lineage-positive cells to apoptotic signaling mediated by the Fas and TNF receptors, which might be useful in improving the efficiency of ex vivo expansion and improving UCB cell engraftment.

Bone Marrow Transplantation (2014) 49, 942–949; doi:10.1038/bmt.2014.79; published online 28 April 2014

INTRODUCTION The identified physiological roles of receptor/ligand interactions in Umbilical cord blood (UCB) is a rich source of hematopoietic the transplant setting include cell interaction with bone-marrow 24 18–22 progenitors with significant implementation in the treatment of stroma and autocrine- and paracrine-trophic signaling. In malignant and nonmalignant disorders.1–6 The use of UCB for addition to these early activities, hematopoietic progenitors fi deficient in Fas and TNF receptors fail to mediate durable- transplantation is limited by the insuf cient cell numbers required 17,18 to attain the threshold of adult recipients and the slow tempo of multilineage reconstitution. engraftment.7–10 Approaches to overcome the small cell numbers Insensitivity of hematopoietic progenitors to apoptosis can be used to improve the outcome of transplants, conferring immune have been implemented with relative success, including trans- 24,25 19–22 plantation of two or more UCB units11,12 and ex vivo expansion of privilege to the graft and fostering progenitor activity. 13–15 A pretransplant apoptotic challenge with Fas-ligand (FasL) and progenitors. The second limitation of UCB transplants is a α fi relatively slow tempo of reconstitution, which is likely an intrinsic TNF- reduces signi cantly the incidence and severity of GvHD in haploidentical- and xenogeneic-murine transplants,26,27 and characteristic of circulating primitive progenitors inexperienced in 20,22 – – improves early myeloid reconstitution. In this study we interaction with the bone-marrow stroma.2 4,7 10 In addition, assessed possible implementation of apoptotic signaling in several although the incidence and intensity of GvHD are generally aspects of UCB cell transplantation. Following documentation of reduced in UCB cell transplants, this reaction might cause 20 fi 16 the insensitivity of UCB progenitors to apoptotic signaling, we signi cant morbidity. assessed the influence of FasL, TNF-α and their combination on We have recently found that the TNF-family receptors have the various lineages as a possible approach to graft preparation different functions in hematopoietic stem and progenitor cells and for transplantation. We reasoned that differential sensitivities of in the differentiated immuno-hematopoietic progeny. In contrast progenitors and differentiated cells to apoptotic signals may be to apoptotic signaling in differentiated cells that mediates used to improve the outcome of ex vivo expansion. We found that homeostatic negative regulation, murine and human precursors fi 17–20 inclusion of FasL in the culture medium improves the ef ciency of are inherently insensitive to apoptosis. Signaling mediated by expansion without impairing progenitor activity. the Fas, TNF and TNF-Related Apoptosis-Inducing Ligand (TRAIL) receptors triggers progenitor activity and synergizes with other inductive factors in promoting hematopoietic differentiation.19,21,22 MATERIALS AND METHODS This mechanism links injury signals, including TNF-family ligands, Cell procurement and isolation with the activation of the hematopoietic system to ensure prompt UCB was obtained from healthy donors pending informed consent recovery from hypoplasia. It also explains the marked upregulation according to the Institutional guidelines of the Beilinson Campus (IRB file 23 of these receptors under conditions of stress hematopoiesis 3656: 920050264). UCB samples were harvested before placental delivery and ubiquitous expression in most primitive progenitors.17,20,21 with citrate-phosphate-dextrose adenine (CPDA-1). Washed-mononuclear

1Department of Pediatric Hematology-Oncology, Frankel Laboratory, Center for Stem Cell Research, Schneider Children's Medical Center of Israel, Petach Tikva, Israel; 2Department of Pediatric Hematology-Oncology, Schneider Children’s Medical Center of Israel, Petach Tikva, Israel; 3Gamida Cell Ltd., Jerusalem, Israel and 4Department of Pediatric Hematology-Oncology, Bone Marrow Transplant Unit, Schneider Children's Medical Center of Israel, Petach Tikva, Israel. Correspondence: Professor N Askenasy, Frankel Laboratory, Center for Stem Cell Research, Schneider Children’s Medical Center of Israel, 14 Kaplan Street, Petach Tikva 49202, Israel. E-mail: [email protected] Received 19 September 2013; revised 27 January 2014; accepted 6 February 2014; published online 28 April 2014 Manipulation of UCB grafts with an apoptotic challenge K Mizrahi et al 943 cells were diluted twofold with medium containing 0.5% human serum IQ Products).19,20,22 Cell cycling was determined with an intracellular dye albumin and were isolated by centrifugation over a Ficoll-density gradient loaded by incubation for 7 min with 5 μM of (5- and 6-)carboxyﬂuorescein (1.077–1.08 g/mL, MP Biomedical, Illkirch, France).19,20 CD34+ progenitors diacetate succinimidyl ester (CFSE; Molecular Probes, Eugene, OR, USA). were immunomagnetically isolated using the CD34 Progenitor Cell Proliferation index was quantiﬁed with the ModFit software (Verity Isolation Kit (Miltenyi Biotec, Bergisch Gladbach, Germany). Software House, Topsham, ME, USA).

Flow cytometry Apoptotic challenge Measurements were performed with a Vantage SE-2 flow cytometer Cell suspensions (5x106 cells/mL) were incubated for variable periods of + (Becton Dickinson, Franklin Lakes, NJ, USA). CD34 cells were identified time at 37°C in 5% CO2 in a humidified incubator in medium, and with using antigen-presenting cells-labeled antibodies (clone 8G12, Pharmin- supplementation of either FLAG-tagged FasL (Axxora, Farmingdale, NY, gen, San Diego, CA, USA) and lineages were quantified using USA) or 20 ng/mL recombinant TNF-α (PeproTech, Rocky Hill, NJ, USA). phycoerythrin-labeled mAb: anti-human CD3 (clone OKT3, Pharmingen), Human cells were incubated in minimum essential medium-α (α-MEM) anti-human CD19 (clone HD37, IQ Products, Groningen, Netherlands) and culture medium (Biological Industries, Beit Haemek, Israel) supplemented anti-human CD33 (clone WM53, IQ Products).27 Expression of the receptors with StemPro Nutrient Supplement (Stem Cell Technologies, Vancouver, was determined using mAb against human Fas (clone DX2, Miltenyi BC, Canada), 2 mML-glutamine and 50 mM 2β-Mercaptoethanol purchased Biotec), TNF-R1 and TNF-R2 (clones 16803 and 22235, respectively, R&D from PeproTech. Dead cells were eliminated by centrifugation over a Ficoll Systems, Abingdon, UK).22 Cell death and apoptosis were determined in gradient (MP Biomedical). cells incubated with 5 μg/mL 7-aminoactinomycin-D (7-AAD, Sigma, St Louis, MO, USA) and Annexin-V (IQ Products), respectively.17 Engraft- ment of UCB cells in NOD.SCID xenochimeras was determined in the bone Ex vivo expansion marrow after 12 weeks using antibodies against murine CD45 (clone Isolated CD34+ cells were cultured in culture bags (American Fluoroseal, 30-F11, eBioscience, San Diego, CA, USA) and human CD45 (clone ML2, Gaithersburg, MD, USA) at 104 cells/mL in α-MEM supplemented with 10%

100 100 medium medium P<0.001 FasL FasL 80 80 TNF-a TNF-a P<0.001 60 60

40 40 % UCB apoptosis % UCB apoptosis 20 20

0 0 CD34 CD3 CD19 CD33 CD34 CD3 CD19 CD33

CD33 P<0.001 2650

2120 MNC CD19 P<0.01 CD33+ CD19+

CD3+ P<0.001 Events CD3 530 TNF-R2 TNF-R1 0 CD34 100 101 102 103 104 Fas Fas 0 20 40 60 80 100 % UCB receptor expression

5 60 P<0.01 50 Medium 40 PI=2.2 4 30 20 10 3 0 050 100150 200 250

2 Events

medium 20 FasL Proliferation index PI=2 1 FasL TNF 10 0 CD34 CD3 CD19 CD33 0 0 50 100 150 200 250 CFSE Figure 1. Expression of death receptors and sensitivity to apoptosis of UCB cells. Fresh UCB samples were incubated for variable periods of time in medium and with 50 ng/mL FasL or 20 ng/mL TNF-α. Apoptosis was assessed by Annexin-V incorporation on gated subsets of CD34+ progenitors, CD3+ T cells, CD19+ B lymphocytes and CD33+ myeloid cells after 24 (a, n = 15–33) and 48 h (b, n = 15–33). (c) Expression of the Fas, TNF-R1 and TNF-R2 receptors in fresh UCB cells (n = 15–33). The right panel presents demonstrative plots of Fas-receptor expression. (d) Proliferation rates as determined from CFSE dilution in the corresponding UCB subsets (n = 9). Right panel is demonstrative of proliferation of CD34+ progenitors in medium and with FasL.

© 2014 Macmillan Publishers Limited Bone Marrow Transplantation (2014) 942 – 949 Manipulation of UCB grafts with an apoptotic challenge K Mizrahi et al 944 fetal bovine serum, 50 ng/mL thrombopoietin, 50 ng/mL interleukin-6, RESULTS 50 ng/mL fms-like tyrosine kinase-3 ligand and 50 ng/mL stem cell factor 28 Differential sensitivity of UCB subsets to receptor-mediated (PeproTech) with 5 mM nicotinamide (Sigma). Cultures were incubated at apoptosis 37 °C in a humidified atmosphere of 5% CO2 for 3 weeks, and were refreshed weekly with an equal volume of fresh medium. Following the observation that hematopoietic progenitors derived from human UCB are resistant to receptor-mediated apoptosis,20 we assessed the sensitivities of lineage-positive UCB subsets to Colony forming unit (CFU) assays these receptors. An apoptotic challenge was applied by supple- 3 3 CFU assays were performed by plating 2.5 × 10 mononuclear cells or 10 mentation of toxic doses of FasL and TNF-α to cells incubated + fi CD34 cells per well in 0.9% methylcellulose-containing Iscove's Modi ed without supporting or stimulating chemokines and measurements Dulbecco Medium (Biological Industries) supplemented with 2 mM were performed by gating on various subsets in mixed-UCB L-glutamine, 30% fetal calf serum, 50 ng/mL stem cell factor, 10 ng/mL + + interleukin-3 and 10 ng/mL rhGM-CSF (PeproTech). Colonies were suspensions. CD34 progenitors and CD3 T cells presented monitored using an inverted Olympus microscope (Center Valley, PA, USA). remarkably low levels of fractional apoptosis, whereas the viability of CD19+ B lymphocytes and CD33+ myeloid cells decreased substantially when incubation was extended from 24 (Figure 1a) Animal model and transplantation to 48 h (Figure 1b). Remarkably, fractional apoptosis appears to be Mice used in this study were C57BL/6 [H2Kb, CD45.2], BALB/c (H2Kd) determined primarily by susceptibilities of the various UCB scid g7 and NOD.CB17-Prkdc (NOD.SCID, H2K ) purchased from Jackson lineages to spontaneous apoptosis with insignificant impact of Laboratories (Bar Harbor, ME, USA) and housed in a barrier facility in the death ligands. This profile questioned whether the various accordance with the guidelines of the Institutional Animal Care and Use UCB lineages express the cognate receptors and are therefore Committee. NOD SCID mice were conditioned with two consecutive daily doses of 25 μg/g BU (Sigma).19,20,22 Levels of human chimerism were responsive to the apoptotic signals. We found that the overall determined in the bone marrow and spleen after 12 weeks using specific patterns of receptor expression were dissociated from sensitivity human markers. Levels of xenochimerism vary widely in individual mice to spontaneous and receptor-induced apoptosis in all UCB grafted with cells from the same sample, therefore this assay was used lineages (Figure 1c). On the one hand, Fas and both TNF receptors primarily to determine possible negative effect of the experimental were equally expressed in ~ 20% of fresh CD34+ progenitors, and procedures on SRC. Fas was expressed in ~ 50% of T cells (Figure 1c), exceeding by far the levels of fractional apoptosis observed in these subsets. In – Statistical analysis addition, the TNF receptors were expressed in 40 60% of the myeloid cells, which displayed progressive apoptosis in liquid Data are presented as means six standard deviations for each experimental protocol. Results in each experimental group were evaluated for culture. Thus expression of the receptor and its activation by the reproducibility by linear regression of duplicate measurements. Differences cognate ligand is not associated with obligatory-apoptotic between the experimental protocols were estimated with a post-hoc signaling. On the other hand, the Fas and TNF receptors were Scheffe t-test and significance was considered at Po0.05. expressed at low levels in on B lymphocytes (Figure 1c), below the P <0.001

CD34 TNF

TNF + FasL P <0.05 Med + FasL Medium+FasL MNC TNF + FasL TNF-α hours 24 48

0 102030405060 % Fas expression

100 5 100 TNF

80 TNF FasL + P<0.01 Med FasL 4 80 Fas + 60 3 60

40 2 40 % apoptosis TNF 20 Proliferation index TNF FasL 20

1 % apoptosis CD34 Med FasL 0 0 MNC D34 0 MNC D34 TNF TNF Med FasL FasL Figure 2. The consequences of TNF-induced Fas expression in UCB cells. (a) UCB cells were incubated in medium and with 20 ng/mL TNF-α for 48 h and 50 ng/mL FasL were supplemented during the second day of culture. (b) Exposure to TNF-α-induced Fas expression in mononuclear cell (n = 26) and CD34+ progenitors (n = 20), irrespective of the presence of FasL. (c) Percent apoptosis of mononuclear cells (n = 26) and CD34+ progenitors (n = 26) under various incubation conditions. (d) Proliferation rates of mononuclear cells and CD34+ progenitors under the corresponding incubation conditions. (e) Apoptosis determined by gating on CD34+ progenitors expressing Fas under various culture conditions (n = 11–14).

TNFα 30 cells 3 P <0.01 20

FasL P <0.05 CFU/10 72 hours 48 hours 10 Medium 24 hours 0 0 10 20 30 40 50 Control Ficoll CFU / 103 UCB cells

100 Medium 80

CD34 60 CD3 CD19 40 α TNF +FasL CD33 % engraftment CD14 20 lin–

0 Medium TNF-α TNF-α +FasL Figure 3. Exposure of UCB cells to death ligands increases the frequency of viable progenitors without impairing SRC activity. (a) Fresh UCB cells were exposed to 20 ng/mL TNFα or 50 ng/mL FasL for 24–72 h, or sequentially to TNF-α and FasL, and dead cells were subsequently removed by centrifugation over a ficoll gradient. Equal numbers of viable cells were plated in semisolid-methylcellulose cultures stimulated with SCF, interleukin-3 and GM-CSF to determine CFU frequency per 1000 viable UCB (n = 7–15). The right panel is demonstrative for progenitor enrichment following elimination of dead cells from 24 h cultures supplemented with TNF-α (n = 15). (b) Engraftment of UCB cells preincubated with TNF-α with and without FasL during the second day as compared with equal number of cells incubated in medium from the same UCB unit. Data are representative of four UCB units grafted into BU-conditioned NOD.SCID mice. (c) Distribution of the CD34+ and lineage-negative (lin-) progenitors, CD3+ T cells, CD19+ B lymphocytes and CD33+ myeloid cells in the bone marrow of NOD.SCID at 12 weeks post-transplantation of UCB cell incubated in medium and sequentially with TNF-α and FasL. levels of fractional apoptosis displayed by this subset. Altogether suggesting that Fas upregulation did not sensitize to apoptosis these data demonstrate that UCB cell death is dominated by the triggered by the cognate ligand. This result was confirmed by the susceptibility to spontaneous apoptosis, which exceeds the reduced FasL-induced fractional apoptosis in the TNF-expanded consequences of apoptotic receptor activation. subset of CD34+Fas+ cells, as compared with higher rates of Next we assessed the cycling rates, as differential proliferation apoptosis in cells incubated in medium (Po0.01, Figure 2e). of viable subsets might affect the measurements of fractional Decreased-fractional apoptosis was evidently caused by upregula- apoptosis within heterogeneous cell populations. Assessment of tion of Fas expression without concurrent sensitization to the cycling rates after 48 h of incubation showed slow prolifera- apoptosis. tion of CD34+ progenitors and CD33+ myeloid cells as compared with T and B cells (Po0.01, Figure 1d), without evident variations induced by the presence of the apoptotic ligands. Thus, low levels Apoptotic signaling increases progenitor frequency and fractional apoptosis of CD3+ T cells were also affected by Exposure of UCB cells to an apoptotic challenge induces apoptosis faster cycling rates, whereas the myeloid subset was most in significant fractions of lineage-positive cells and spares CD34+ sensitive to spontaneous apoptosis and presented the slowest progenitors. To determine potential detrimental effects on rates of proliferation. progenitor function, the frequency of CFU was assessed in semisolid-methylcellulose cultures after joint exposure to FasL and TNF-α. Dead cells were removed by centrifugation over a ficoll TNF-α upregulates Fas expression without sensitizing to apoptosis gradient and equal numbers of viable cells were plated in Besides direct activity of TNF-α through its cognate receptors, this semisolid cultures stimulated with SCF, interleukin-3 and GM-CSF. cytokine has been also shown to operate through induction of Fas The number of myeloid colonies increased after UCB exposure to expression.29,30 To evaluate this indirect mechanism, UCB cells toxic concentrations of FasL and TNF-α for 24–48 h, emphasizing were exposed to TNF-α and were sequentially submitted to an enrichment of apoptosis-insensitive progenitors following apoptotic challenge with FasL during the second day of culture elimination of dead cells (Figure 3a). Increased-CFU frequencies (Figure 2a). As expected, short-term incubation with TNF-α- were also observed after sequential TNF-mediated upregulation of induced expression of Fas in UCB cells, particularly in ~ 40% of Fas and cross-linking by FasL consistent with disproportional CD34+ progenitors (Figure 2b). However, both fractional apoptosis apoptosis of lineage-positive UCB cells. However, the ligands did (Figure 2c) and proliferation rates (Figure 2d) were unaffected, not synergize in induction of apoptosis in sensitive cells, indicating

© 2014 Macmillan Publishers Limited Bone Marrow Transplantation (2014) 942 – 949 Manipulation of UCB grafts with an apoptotic challenge K Mizrahi et al 946 redundant activity. Progenitor death and dysfunction after Exposure to death ligands improves the outcome of ex vivo UCB extended culture for 72 h has long been recognized to decrease expansion the number of colonies, irrespective of the presence of apoptotic One of the proposed solutions to the relatively low numbers of 31–33 19 signals. Like TRAIL, TNF-α had a mild-protective effect (non- progenitors in UCB is ex vivo expansion,13–15 a process that would significant) on UCB cells incubated for 72 h. Altogether these data be best implemented by induction of proliferation and arrest of demonstrate functional-negative selection of progenitors by progenitor differentiation. Since supportive chemokines induce induction of receptor-mediated apoptosis in sensitive cells, progenitor differentiation in these extended cultures,34,35 we despite no apparent differences in viability in direct measure- reasoned that differential sensitivity to apoptosis might be useful ments of apoptosis (Figures 1a, b and 2c). for depletion of the differentiated progeny. We used a CD34+ expansion protocol based on nicotinamide28 and supplemented toxic doses of FasL (50 ng/mL) during the third week of culture, Apoptotic signaling does no impair SRC activity the period associated with largest quantitative expansion of cells. The most significant prerequisite of expansion protocols is to Fas cross-linking during the terminal culture period increased the maintain the activity of repopulating cells, which is best evaluated fractions of CD34+ progenitors exceeding the expansion achieved in a surrogate assay of SRC representing a more primitive subset by this protocol (Figure 4a). of progenitors than CFU.19 We have demonstrated that exposure To determine the quality of progenitor exposure to FasL during to FasL20 and TNF-α22 for 48 h does not impair SRC activity, and the expansion culture, cells were assessed in two functional subsequently assessed the impact of sequential exposure to both assays. Increased CD34+ fractions following expansion in the ligands. Transplantation of equal numbers of viable cells following presence of FasL might be mediated either by arrest of sequential incubation with TNF-α and FasL into NOD.SCID mice differentiation or induction of apoptosis in cells generated by showed comparable levels of xenochimerism as cells incubated spontaneous differentiation. To determine the mechanism of with and without TNF-α alone (Figure 3b). Furthermore, the bone enrichment, the expanded cells were first assayed in semisolid- marrow of NOD.SCID xenochimeras presented increased fractions methylcellulose cultures. Plating of equal number of total of human B lymphocytes after 12 weeks, which is the dominant expanded cells showed a significant increase in myeloid lineage developing from UCB progenitors (Figure 3c). Altogether progenitor activity following exposure to FasL (Po0.01, these data demonstrate that pretransplant TNF-mediated Figure 4b), indicative of a net increase in CFU frequency following upregulation of Fas and cross-linking of this receptor does not Fas cross-linking. In a second assay the dead cells were eliminated suppress but rather stimulates SRC activity in vivo. by centrifugation over a ficoll gradient (Figure 3a), and equal

100 P<0.001 P 80 <0.01 Ficoll cells P 3 P<0.001 60 <0.001 / 10 +

40 P <0.01 Control CD34 20 Expanded FasL Expanded 0 Fresh Medium FasL 0 102030405060 3 Expanded CFU/10 expanded cells

4 50 10 Bone marrow Medium 3 Spleen 10 40 102

101 38 30 100 100 101 102 103 104 104 FasL

20 Mouse CD45 103 10 102 % human chimerism 101 39

0 100 100 101 102 103 104 Medium FasL Human CD45 Figure 4. Death ligands increase progenitor frequency in ex vivo expanded-UCB progenitors. CD34+ progenitors were immunomagnetically isolated and expanded using a nicotinamide-based protocol for 3 weeks in liquid culture. (a) Absolute numbers of expanded CD34+ progenitors (per 103 total-nucleated cells) were increased by the presence of 50 ng/mL FasL during the third week of culture (n = 3). (b) Equal number of total-nucleated cells (control) following expansion with and without exposure to FasL during the third week were plated in semisolid-methylcellulose cultures stimulated with SCF, interleukin-3 and GM-CSF. CFU frequency is given per 1000 viable cells (n = 3). Subsequently equal numbers of viable cells were plated after removal of dead cells over ﬁcoll gradient. (c) Equal numbers of viable cells following expansion with and without FasL were grafted into NOD.SCID mice conditioned with two doses of 25 μg/g BU, and engraftment was assessed in the bone marrow and spleen after 12 weeks. The right panel presents demonstrative plots of human cell engraftment in the bone marrow following expansion in medium and with FasL (representative of two independent expansion cultures).

Bone Marrow Transplantation (2014) 942 – 949 © 2014 Macmillan Publishers Limited Manipulation of UCB grafts with an apoptotic challenge K Mizrahi et al 947 numbers of viable cells were plated in semisolid cultures numeric expansion,13 increased substantially the fractions of (Figure 4b). Elimination of the dead cells resulted in marked CD34+ progenitors. This was achieved primarily through induction enrichment in CFU-GM frequency under both expanded and FasL- of receptor-mediated apoptosis in the differentiated progeny, supplemented cultures (Po0.001), consistent with progenitor conferring a competitive advantage to expansion of uncommitted enrichment in fresh UCB samples. The proportional enrichment in progenitors. Expansion was initiated with isolated CD34+ progeni- net numbers and progenitor fractions mediated by FasL suggest tors that are largely insensitive to apoptosis, and spontaneous that the main mechanism was induction of apoptosis in sensitive differentiation during the extended culture caused gradual cells during terminal stages of the expansion procedure. In the sensitization to Fas-receptor signaling and consequently negative second assay, the quality of expanded progenitors was assessed regulation by apoptosis.18–22 It was therefore evident that the in xenogeneic transplants in BU-conditioned NOD.SCID mice. dominant mechanism of increased-CFU percentage in fresh- and Progenitors expanded with and without Fas cross-linking showed expanded-UCB cells was owing to the depletion of more comparable levels of engraftment in the bone marrow and spleen differentiated cells, which are largely insignificant and dispensable after 12 weeks (Figure 4c), demonstrating that exposure to FasL in in the process of hematopoietic reconstitution. terminal stages of expansion did not impair SRC activity. The second relative drawback of UCB grafts is a slower pace of engraftment of primitive progenitors inexperienced in interaction with the bone-marrow stroma.7–10 Proposed solutions include DISCUSSION direct inoculation of UCB cells into the bone marrow to overcome Our findings extend the evidence of insensitivity of UCB-derived ineffective UCB cell homing,12,38,39 and increasing the effective hematopoietic progenitors to apoptosis mediated by the Fas and number through the combination of several cord units11,12 and TNF receptors, in addition to resistance to apoptotic signaling ex vivo expansion of progenitors.13–15 In addition to generation of mediated by the TRAIL receptors.19,20,22 Resistance of human a progenitor-enriched graft by functional-negative selection (as UCB progenitors to receptor-mediated apoptosis, consistently compared with phenotypic identification), pretransplant exposure observed in mobilized-peripheral blood cells22,27 and murine- to the death ligands enhanced progenitor activity. Qualitative SRC bone marrow,17–19,21 evolves as an inherent characteristic of activity in vivo and clonogenic assays of human hematopoietic hematopoietic precursors.20 Furthermore, we demonstrate by progenitors in vitro have demonstrated trophic signaling direct measurements of viability and functional hematopoietic- mediated by TRAIL-R1,19 Fas20 and TNF-R1.22 Here we document reconstitution assays that TNF-induced Fas expression in 40% of that joint exposure of progenitors to Fas and TNF-α also enhanced the CD34+ progenitors does not sensitize to apoptosis induced by SRC activity promoting the generation of B lymphocytes, which receptor cross-linking, similar to observations in murine-bone are the first lineage to develop from human progenitors in mouse marrow.18 Hematopoietic cells progressively acquire sensitivity to xenochimeras. Thus, ex vivo exposure of UCB progenitors to receptor-mediated apoptosis in the course of differentiation, death ligands and their combination simultaneously serves two which becomes the dominant mechanism of physiological- functions: negative selection of progenitors and activation of their negative regulation of the mature progeny.18–22 Differential function. sensitivities of progenitors and differentiated cells to receptor- A third envisioned feature with the potential to improve the mediated apoptosis can be employed to improve the outcome of outcome of human hematopoietic-cell engraftment is GvHD UCB transplants. prophylaxis. Although GvH reactivity is undetected in UCB Negative selection by induction of receptor-mediated apoptosis xenochimeras and the severity of GvHD is generally reduced in mature lineages generates a larger graft than positive selection in human UCB transplants, it sometimes imposes severe of progenitors using identifying markers. Considering the infidelity morbidity.7–10,16 We have recently demonstrated that exposure of CD34 as an identifier of hematopoietic progenitors,36 this of donor inoculum to an apoptotic challenge in the absence of approach includes more cells with hematopoietic-reconstituting host-specific sensitization reduces GvHD severity in mice grafted potential in the graft. In fresh UCB samples, B lymphocytes and with haploidentical T cells26 and human mobilized-peripheral myeloid cells are sensitive to spontaneous apoptosis during blood xenografts.27 In view of the preserved-engraftment 24–48 h of incubation and their death increases the percentage of facilitating and graft versus tumor activities of T cells pretreated progenitors. Despite insignificant variations in the measured levels with death ligands, it remains to be determined whether UCB of fractional apoptosis, more sensitive functional assays revealed T cells have reduced-GvH reactivity following functional depletion. increased-progenitor frequencies. Consistent with application of T cells in UCB were largely insensitive to receptor-mediated each one of the death ligands alone, combined exposure to FasL apoptosis, as immature cells40–42 with reduced-cytotoxic activity in and TNF-α has no detrimental consequences, in contrast to response to cytokine stimulation in vitro43 that become sensitive negative effects in chemokine-supplemented cultures29,30 that to negative regulation only after activation.44–46 induce cell cycling and differentiation.34,35 Both TNF receptors Hematopoietic progenitors derived from different sources equally contribute to upregulation of Fas in UCB progenitors display variable susceptibility to spontaneous apoptosis and within several events mediated by receptor crosstalk.37 Notably, differential responses to activation of the TNF-family receptors. joint exposure to TNF-α and FasL does not result in cumulative UCB cells are safely maintained in liquid suspensions for 24–48 h, enrichment in progenitors, suggesting that nonprogenitor cells whereas in mobilized- peripheral blood cells similar rates of share redundant sensitivities to both receptors. apoptosis are observed after 4–12 h of incubation.27 The relatively Pretransplant activation of the Fas and TNF receptors in UCB activated state of mobilized-peripheral blood cells is also cells is potentially beneficial in overcoming the relative limitations recognized in the shorter period required to activate progenitor of these grafts. The first drawback is the low absolute number of function by receptor-mediated trophic signals22 and the remark- progenitors in UCB, which are often insufficient to attain able difference in viability of T cells in culture.27 It is difficult to threshold-donor inoculum for transplantation in adult recipients. predict the potential benefit of application of the current Differential sensitivities of progenitors and differentiated cells to approach to overcome the particular limitations of UCB cells, apoptosis were used to increase the yield of an ex vivo expansion and whether it can reduce the 10–15 day delay in UCB cell protocol,28 which would be optimally achieved by induced engraftment as compared with bone marrow and mobilized- proliferation of uncommitted progenitors and concomitant peripheral blood cells.47,48 Various approaches to foster engraft- inhibition of their commitment to differentiate. Application of ment are being considered48 to reduce the period of neutropenia the apoptotic challenge during the terminal period of culture, following UCB transplants, in particular for treatment of associated with the highest rates of active proliferation and malignancies.47,49,50 Enhancing myeloid reconstitution, particularly

© 2014 Macmillan Publishers Limited Bone Marrow Transplantation (2014) 942 – 949 Manipulation of UCB grafts with an apoptotic challenge K Mizrahi et al 948 in early post-transplant stages may reduce the threat of failure of marrow transplant from an HLA-identical sibling. New Engl J Med 2000; 342: engraftment51 and alleviate the high incidence of infections 1846–1854. during the period of neutropenia.52 17 Pearl-Yafe M, Yolcu ES, Stein J, Kaplan O, Shirwan H, Yaniv I, Askenasy N. In summary, resistance of UCB-derived hematopoietic Expression of Fas and Fas-ligand in donor hematopoietic stem and progenitor cells is dissociated from the sensitivity to apoptosis. Exp Hematol 2007; 35: progenitors to apoptosis can be harnessed to improve the – outcome of transplants. Polarized sensitivities of UCB cells to 1601 1612. 18 Pearl-Yafe M, Mizrahi K, Stein J, Stein J, Yolcu ES, Kaplan O et al. Tumor necrosis spontaneous and receptor-mediated apoptosis can be used for factor receptors support murine hematopoietic progenitor function in the early graft preparation and enrichment of progenitors by negative stages of engraftment. Stem Cells 2010; 28: 1270–1280. selection, with concomitant-progenitor activation and reduction 19 Mizrahi K, Stein J, Pearl-Yafe M, Kaplan O, Yaniv I, Askenasy N. Regulatory func- of GvHD severity. These distinct mechanisms are expected to tions of TRAIL in hematopoietic progenitors: human umbilical cord blood and converge and improve the relative drawbacks of UCB grafts. murine bone marrow transplantation. Leukemia 2010; 24:1325–1334. 20 Mizrahi K, Stein J, Kaplan O, Yaniv I, Zipori D, Askenasy N. Resistance of hematopoietic progenitors to Fas-mediated apoptosis is actively sustained by NFκB CONFLICT OF INTEREST with a characteristic transcriptional signature. Stem Cell Dev 2013; 23:676–686. 21 Pearl-Yafe M, Stein J, Yolcu ES, Farkas DL, Shirwan H, Yaniv I et al. Fas transduces Dr Tony Peled has ownership interest in Gamida Cell. The other authors declare no dual apoptotic and trophic signals in hematopoietic progenitors. Stem Cells 2007; conflict of interest. 25: 3194–3203. 22 Mizrahi K, Stein J, Yaniv I, Kaplan O, Askenasy N. TNF-α has tropic rather than ACKNOWLEDGEMENTS apoptotic activity in human hematopoietic progenitors: involvement of TNF receptor-1 and caspase-8. Stem Cells 2013; 31:156–166. This work was funded by grants from the Frankel Trust for Experimental Bone Marrow 23 Saheki K, Fujimori Y, Takemoto Y, Kakishita E. Increased expression of Fas (APO-1, Transplantation. We thank Mrs Ela Zuzovsky and Mrs Ana Zemliansky for their CD95) on CD34+ haematopoietic progenitor cells after allogeneic bone marrow outstanding technical support. transplantation. Br J Haematol 2000; 109:447–452. 24 Pearl-Yafe M, Yolcu ES, Stein J, Kaplan O, Yaniv I, Shirwan H et al. Fas ligand enhances hematopoietic cell engraftment through abrogation of alloimmune REFERENCES responses and nonimmunogenic interactions. Stem Cells 2007; 25: 1448–1455. 1 Fernández MN, Millán I, Gluckman E. Cord-blood transplants. New Engl J Med 25 Askenasy EM, Shushlav Y, Sun Z, Shirwan H, Yolcu ES, Askenasy N. Engineering of fi 1999; 340: 1287–1288. bone marrow cells with fas-ligand protein-enhances donor-speci c tolerance to 43 – 2 Rosler ES, Brandt JE, Chute J, Hoffman R. An in vivo competitive repopulation solid organs. Transplant Proc 2011; : 3545 3548. assay for various sources of human hematopoietic stem cells. Blood 2000; 96: 26 Askenasy N, Ash S, Yaniv I, Stein J. Depletion of naïve lymphocytes with Fas ligand 3414–3421. ex vivo prevents graft-versus-host disease without impairing T cell support of 19 3 Tanavde VM, Malehorn MT, Lumkul R, Gao Z, Wingard J, Garrett ES et al. Human engraftment or graft-versus-tumor activity. Biol Blood Marrow Transplant 2013; : stem-progenitor cells from neonatal cord blood have greater hematopoietic 185–195. expansion capacity than those from mobilized adult blood. Exp Hematol 2002; 30: 27 Mizrahi K, Yaniv I, Ash S, Stein J, Askenasy N. Apoptotic signaling through Fas and 816–823. TNF receptors ameliorates GVHD in mobilized peripheral blood grafts. Bone Marrow 4 Frassoni F, Podesta M, Maccario R, Giorgiani G, Rossi G, Zecca M et al. Cord blood Transplant 2014; 49:640–648. transplantation provides better reconstitution of hematopoietic reservoir 28 Peled T, Shoham H, Aschengrau D, Yackoubov D, Frei G, Rosenheimer GN et al. compared with bone marrow transplantation. Blood 2003; 102: 1138–1141. Nicotinamide, a SIRT1 inhibitor, inhibits differentiation and facilitates expansion of 5 Rocha V, Labopin M, Sanz G, Arcese W, Schwerdtfeger R, Bosi A et al. Transplants hematopoietic progenitor cells with enhanced bone marrow homing and of umbilical-cord blood or bone marrow from unrelated donors in adults with engraftment. Exp Hematol 2012; 40: 342–355. acute leukemia. New Engl J Med 2004; 351: 2276–2285. 29 Maciejewski J, Selleri C, Anderson S, Young NS. Fas antigen expression on CD34+ 6 Lekakis L, Giralt S, Couriel D, Shpall EJ, Hosing C, Khouri IF et al. Phase II study of human marrow cells is induced by interferon gamma and tumor necrosis factor unrelated cord blood transplantation for adults with high-risk hematologic alpha and potentiates cytokine-mediated hematopoietic suppression in vitro. malignancies. Bone Marrow Transplant 2006; 38:421–426. Blood 1995; 85: 3183–3190. 7 Tse W, Laughlin M. Cord blood transplantation in adult patients. Cytotherapy 2005; 30 Bryder D, Ramsfjell V, Dybedal I, Theilgaard-Mönch K, Högerkorp CM, Adolfsson J 7:228–242. et al. Self-renewal of multipotent long-term repopulating hematopoietic stem 8 Schoemans H, Theunissen K, Maertens J, Boogaerts M, Verfaillie C, Wagner J. Adult cells is negatively regulated by Fas and tumor necrosis factor receptor activation. umbilical cord blood transplantation: a comprehensive review. Bone Marrow J Exp Med 2001; 194:941–952. Transplant 2006; 38:83–93. 31 Guenechea G, Segovia JC, Albella B, Lamana M, Ramírez M, Regidor C et al. 9 Brown JA, Boussiotis VA. Umbilical cord blood transplantation: basic biology and Delayed engraftment of nonobese diabetic severe combined immunodeficient clinical challenges to immune reconstitution. Clin Immunol 2008; 127:286–297. mice transplanted with ex vivo–expanded human CD34 cord blood cells. Blood 10 Rocha V, Broxmeyer H. New approaches for improving engraftment after Cord 1999; 93: 1097–1105. Blood Transplantation. Biol Blood Marrow Transplant 2010; 16(Suppl 1): 32 Glimm H, Oh IH, Eaves CJ. Human hematopoietic stem cells stimulated to pro- S126–S132. liferate in vitro lose engraftment potential during their S/G(2)/ M transit and do 11 Barker JN, Weisdorf DJ, DeFor TE, Barker JN, Weisdorf DJ, DeFor TE et al. not reenter G(0). Blood 2000; 96: 4185–4193. Transplantation of 2 partially HLA-matched umbilical cord blood units to enhance 33 Ratajczak J, Kucia M, Reca R, Zhang J, Machalinski B, Ratajczak MZ. Quiescent engraftment in adults with hematologic malignancy. Blood 2005; 105: 1343–1347. CD34+ early erythroid progenitors are resistant to several erythropoietic 12 Brunstein CG, Barker JN, Weisdorf DJ, Defor TE, McKenna D, Chong SY et al. 'inhibitory' cytokines; role of FLIP. Br J Haematol 2003; 123:160–169. Intra-BM injection to enhance engraftment after myeloablative umbilical cord 34 Takenaka K, Nagafuji K, Harada M, Mizuno S, Miyamoto T, Makino S et al. In vitro blood transplantation with two partially HLA-matched units. Bone Marrow expansion of hematopoietic progenitor cells induces functional expression of Fas Transplant 2009; 43:935–940. antigen (CD95). Blood 1996; 88: 2871–2877. 13 Peled T, Landau E, Mandel J, Glukhman E, Goudsmid NR, Nagler A et al. Linear 35 Santiago-Schwarz F, Borrero M, Tucci J, Palaia T, Carsons SE. In vitro expansion of polyamine copper chelator tetraethylenepentamine augments long-term ex vivo CD13+CD33+ dendritic cell precursors from multipotent progenitors is regulated expansion of cord blood-derived CD34+ cells and increases their engraftment by a discrete fas-mediated apoptotic schedule. J Leukoc Biol 1997; 62: 493–502. potential in NOD/SCID mice. Exp Hematol 2004; 32:547–555. 36 Askenasy N, Yaniv I, Stein J, Sharkis SJ. Our perception of developmental plasticity: 14 Araki H, Mahmud N, Milhem M, Nunez R, Xu M, Beam CA et al. Expansion esse est percipi (to be is to be perceived)? Curr Stem Cell Res Ther 2006; 1:85–94. of human umbilical cord blood SCID-repopulating cells using chromatin- 37 Mizrahi K, Askenasy N. Activation and crosstalk between TNF family receptors in modifying agents. Exp Hematol 2006; 34:140–149. umbilical cord blood cells is not responsible for loss of engraftment capacity 15 de Lima M, McMannis J, Gee A, Komanduri K, Couriel D, Andersson BS et al. following culture. Am J Stem Cells 2013; 2:155–164. Transplantation of ex vivo expanded cord blood cells using the copper chelator 38 Frassoni F, Gualandi F, Podestà M, Raiola AM, Ibatici A, Piaggio G et al. Direct tetraethylenepentamine: a phase I/II clinical trial. Bone Marrow Transplant 2008; intrabone transplant of unrelated cord-blood cells in acute leukaemia: a phase I/ 41: 771–778. II study. Lancet Oncol 2008; 9: 831–839. 16 Rocha V, Wagner JE Jr, Sobocinski KA, Klein JP, Zhang MJ, Horowitz MM et al. 39 Rocha V, Labopin M, Ruggeri A, Podestà M, Gallamini A, Bonifazi F et al. Unrelated Graft-versus-host disease in children who have received a cord-blood or bone cord blood transplantation: outcomes after single-unit intrabone injection

Bone Marrow Transplantation (2014) 942 – 949 © 2014 Macmillan Publishers Limited Manipulation of UCB grafts with an apoptotic challenge K Mizrahi et al 949 compared with double-unit intravenous injection in patients with hematological 47 Rocha V, Cornish J, Sievers EL, Filipovich A, Locatelli F, Peters C et al. Comparison malignancies. Transplantation 2013; 95: 1284–1291. of outcomes of unrelated bone marrow and umbilical cord blood transplants in 40 Harris DT, LoCascio J, Besencon FJ. Analysis of the alloreactive capacity of human children with acute leukemia. Blood 2001; 97: 2962–2971. umbilical cord blood: implications for graft-versus-host disease. Bone Marrow 48 Rocha V, Broxmeyer HE. New approaches for improving engraftment after cord Transplant 1994; 14:545–553. blood transplantation. Biol Blood Marrow Transplant 2010; 16(Suppl. 1): 41 Aggarwal S, Gupta A, Nagata S, Gupta S. Programmed cell death (apoptosis) in S126–S132. cord blood lymphocytes. J Clin Immunol 1997; 17:63–73. 49 Gluckman E, Rocha V, Boyer-Chammard A, Locatelli F, Arcese W, Pasquini I et al. 42 Sato K, Nagayama H, Takahashi TA. Aberrant CD3- and CD28-mediated signaling Outcome of cord-blood transplantation from related and unrelated donors. events in cord blood T cells are associated with dysfunctional regulation of Fas Eurocord Transplant Group and the European Blood and Marrow Transplantation ligand-mediated cytotoxicity. J Immunol 1999; 162: 4464–4471. Group. New Engl J Med 1997; 337: 373–381. 43 Risdon G, Gaddy J, Stehman FB, Broxmeyer HE. Proliferative and cytotoxic 50 Laughlin MJ, Barker J, Bambach B, Koc ON, Rizzieri DA, Wagner JE et al. responses of human cord blood T lymphocytes following allogeneic stimulation. Hematopoietic engraftment and survival in adult recipients of umbilical- Cell Immunol 1994; 154:14–24. cord blood from unrelated donors. New Engl J Med 2001; 344: 44 El Ghalbzouri A, Drennou B, Blancheteau V. An in vitro model of allogeneic 1815–1822. stimulation of cord blood: induction of Fas independent apoptosis. Hum Immunol 51 Moscardó F, Sanz J, Senent L, Cantero S, de la Rubia J, Montesinos P et al. Impact 1999; 60: 598–607. of hematopoietic chimerism at day +14 on engraftment after unrelated donor 45 Lin SJ, Cheng PJ, Hsiao SS, Lin HH, Kuo ML. Differential effect of IL-15 and IL-2 on umbilical cord blood transplantation for hematologic malignancies. Haematolo- survival of PHA-activated umbilical cord blood T cells. Am J Hematol 2005; 80: gica 2009; 94: 827–832. 106–112. 52 Hamza NS, Lisgaris M, Yadavalli G, Nadeau L, Fox R, Fu P et al. Kinetics of myeloid 46 Lin SJ, Lee CC, Cheng PJ, See LC, Kuo ML. Susceptibility to Fas and TNF-α receptor and lymphocyte recovery and infectious complications after unrelated umbilical mediated apoptosis of anti-CD3/anti-CD28-activated umbilical cord blood T cells. cord blood versus HLA-matched unrelated donor allogeneic transplantation Pediatr Allergy Immunol 2000; 20:392–398. in adults. Br J Haematol 2004; 124: 488–498.