Brief ex vivo Fas-ligand incubation attenuates GvHD without compromising stem cell graft performance

Graft versus host disease (GvHD) remains a limiting factor for successful hematopoietic stem cell transplantation (HSCT). T cells and antigen-presenting cells (APCs) are major components of the hematopoietic G-CSF mobilized peripheral blood cell (MPBC) graft. Here we show that a short incubation (2 h) of MPBCs with hexameric Fas ligand (FasL) selectively induces apoptosis of specific donor T cell subsets and APCs but not of CD34+ cells. FasL treatment preferentially induces apoptosis in mature T cell subsets which express high levels of Fas (CD95), such as T stem cell memory, T central memory, and T effector memory cells, as well as TH1 and TH17 cells. Anti-CD3/CD28 stimulated T cells derived from FasL-treated-MPBCs express lower levels of CD25 and secrete lower levels of IFN-γ as compared to control cells not treated with FasL. FasL treatment also induces apoptosis of transitional, naïve, memory and plasmablastoid B cells leading to a reduction in their numbers in the graft and following engraftment in transplanted mice. Most importantly, ex vivo treatment of MPBCs with FasL prior to transplant in conditioned NOD-scid IL2Rγnull (NSG) mice prevented GvHD while preserving graft versus leukemia (GvL) effects, and leading to robust stem cell engraftment.


Introduction
Allogeneic hematopoietic stem-cell transplantation (HSCT) can cure a wide range of malignant and non-malignant hematological diseases. Following transplantation of G-CSF-mobilized peripheral blood cells (MPBCs) [1], T cells that are contained in the graft promote hematopoietic engraftment, T-cell immunity and potent graft versus leukemia/lymphoma effects (GvL) [2][3][4]. A double-edged sword, these T cell can also mediate graft-versus-host disease (GvHD) when they recognize and respond to host allo-antigens presented by host or donor antigen-presenting cells (APCs) [5,6].
A variety of in vivo or ex vivo modalities are available to prevent GvHD, but these same agents and procedures may attenuate immune reconstitution, increase the risk of infections and/or abrogate T-cell-mediated GvL effects [7,8].
It has been shown in both preclinical and clinical models that specific T cell subsets exert positive and negative control over the GvHD reaction. In mice, infusion of naïve T cells (CD45RA + ) induces severe acute GvHD (aGvHD). By contrast, infusion of allogeneic T central memory (CM) induces milder GvHD, and T effector memory cell infusion (EM) does not cause significant GvHD [9][10][11][12][13][14][15]. Surprisingly, however, in a single-arm clinical trial in which CD45RA + cells were depleted from allografts, the incidence of aGvHD was not reduced [16,17].
Fas and its cognate ligand are important regulators of diverse cellular responses. Binding of Fas ligand (FasL; CD178) to Fas (receptor; CD95) initiates a cascade of events culminating in apoptosis [18]. Under physiological conditions, this process curbs exuberant immune responses [19][20][21]. In the transplant setting, soluble FasL efficiently induces apoptosis in donor anti-host T cells that have been primed by co-culture with irradiated host cells (the mixed lymphocyte reactions (MLR)). Infusion of MLR-primed lymphocytes followed by ex vivo FasL treatment prevents GvHD in preclinical models while preserving antitumor reactivity [22]. Similar results were obtained by ex vivo FasL/MLR depletion of allo-reacting human donor anti-host T cells [23]. Askenasy et al. demonstrated that ex vivo incubation of recipient-naïve donor mouse lymphocytes with FasL causes apoptosis of activated CD8 and CD4 cells with a concurrent increase in the regulatory T-cell (Treg) population [24]. Importantly, these authors showed that treatment of murine recipient-naïve haploidentical splenocytes with FasL in the context of F1 to hybrid transplantation led to lower clinical and histological GvHD scores of skin and gastrointestinal tract as compared with the fatal GvHD seen following the infusion of control splenocytes. Surprisingly, incubation of allografts with FasL (a protein usually associated with apoptosis and cell death) actually facilitated hematopoietic progenitor engraftment and showed trophic effects on primitive hematopoietic precursors. Mice so engrafted demonstrated preserved and potent graft-versustumor reactions [24].
As a prelude to the clinical use of FasL to prevent GvHD in patients undergoing stem cell transplantation, we have explored the effects of a brief (2 h) incubation of MPBCs derived from healthy, adult donors with FasL on graft composition (T cell subsets, antigen-presenting cells, in vitro hematopoietic colony formation and CD34 + hematopoietic progenitor cells). FasL-treated MPBC's were infused in immune-deficient mice (both tumor-and non-tumor-bearing) to assess the impact of FasL exposure on the development of both GvHD and GvL. We demonstrate that ex vivo treatment of MPBCs with FasL prior to infusion in NSG mice prevented GvHD in this xenograft setting while preserving both robust stem cell engraftment and GvL.

Materials and methods
Apheresis sample collection and FasL treatment G-CSF mobilized peripheral blood cells (MPBCs) were collected by apheresis from healthy donors at the Schneider Children Medical Center, (Institutional Review Board [IRB] approval No. 0613-14-RMC) or at Rambam Medical Center, (IRB approval No. 506-14-RMB).

Colony-forming unit assay
In vitro colony-forming assays were used as surrogate measures of graft potency. Mononuclear fraction was separated over lymphocyte separation medium (LSM, MP Biomedicals, Santa Ana, California, US). Cells were diluted with IMDM + 2% FBS and MethoCult™ H4034 medium, seeded in a 6 well SmartDish plate (all reagents purchased from StemCell Technologies, Vancouver, BC, Canada) and incubated at 37°C, in 5% CO 2 , with ≥ 95% humidity for 12-16 days. The number of colonies was expressed per 1 × 10 4 seeded total nucleated cells (TNCs).

Flow cytometry
Red blood cells from MPBCs or mouse spleen, blood or BM cell suspensions were lysed using ACK lysis buffer (A10492-01, Thermo Fisher Scientific, Waltham MA, USA) and 1 × 10 6 cells were stained with various combinations of fluorescence-conjugated anti-human antibodies (accept for CD45 that was of either mouse or human origin as indicated in the text). For cell surface staining, cells were incubated with antibodies for 15 min at 4°C in the dark. Intracellular staining was performed using either Inside Stain Kit or FoxP3 Staining Buffer Set (130-090-477 or 130-093-142, respectively, Miltenyi Biotech, Bergisch Gladbach, Germany) according to manufacturer's instructions.

Xenogeneic GvHD and engraftment models
NOD-scid IL2Rgamma-null (NSG) mice (Jackson Laboratory, Bar Harbor MN, USA) were housed in a pathogen-free facility and handled in accordance with the guidelines of the Animal Care and Use Committee of the Rabin Medical Center, Petach Tikva, Israel. For each experiment 7-10 (acceptable group size based on statistical and ethical considerations) female NSG mice (7-9 weeks) were irradiated with 2 or 2.75 Gy (CLINAC-DBX linear accelerators, Varian Medical System, Palo Alto, CA, USA) 24-48 h prior to infusion of ⁓5 × 10 6 unfractionated MPBCs (total nucleated cell (TNC)) or 0.1 × 10 6 purified CD34 + cells/mouse, respectively. Mice were randomized according to age and body weight. Monitoring of GvHD clinical score was carried out twice weekly according to the murine clinical grading system described by Cooke et al. [27]. Weight loss, hunched posture, skin lesions, dull fur, and mobility were each assigned scores of 0 (absent), 1 (moderate) or 2 (severe). The investigator was blinded to the group allocation during each experiment.
Mice were sacrificed in case of weight loss ≥20% of initial weight or upon reaching a clinical GvHD score of ≥7. At each study termination endpoint; spleen, blood and BM were harvested, and human cell engraftment was analyzed by flow cytometry. Mice that exhibited <1% hCD45 in their bone marrow or spleen and <0.1% hCD45 in their blood were considered as not engrafted and were excluded from mean calculation.
During the respective scheduled study termination, spleen weight was recorded. The liver of each animal was collected and fixed for at least 48 h in 10% natural buffered formalin (approximately 4% paraformaldehyde) and H&E staining of Paraffin sections was performed. Histopathological changes were scored using semi-quantitative grading

In vivo GvHD versus leukemia models
Mice were housed in a pathogen-free facility and handled in accordance with the guidelines of the Animal Care and Use Committee of the Hadassah Medical Center, Jerusalem, Israel. Mice were irradiated with 2 Gy, and MV4-11 (10 × 10 6 cells/mouse, ATCC CRL-9591, Manassas, VA, USA were infused intravenously (IV) within 24 h. After 4-6 h, mice were infused IV with either FasL-treated-MPBCs, control MPBCs or vehicle. Monitoring of GvHD clinical score was carried out twice weekly until 3 weeks post-transplantation, when the mice were sacrificed; blood, BM and spleens were harvested. The leukemic burden and the degree of human leukocyte engraftment were assessed using flow cytometry (MACSQuant Analyzer 10 and MACSQuantify Software 2.11, Miltenyi Biotech, Bergisch Gladbach, Germany). The percentage and relative number of the leukemic cells (hCD45 + hCD123 + ), human hematopoietic cells (hCD45 + hCD123 − ) and murine hematopoietic cells (mCD45 + hCD45 -hCD123 − ) were evaluated (all antibodies manufactured by Miltenyi Biotech, Bergisch Gladbach, Germany).

Statistical analysis
For every figure, statistical tests are indicated.
ANOVA parametric test or two tailed paired T-test for comparison MPBC vs MPBC + FasL samples, data presented as average of three technical replicates for each of at least three independent experiments each performed using individual MPBC donation.
Mann-Whitney test and Log-rank (Mantel-Cox) test for animal studies, at least seven animals/ group was tested.
Two tailed unpaired Student's t test was applied for technical triplicates of individual representative tests. GraphPad Prism version 8.0 (San Diego, CA USA) was used for statistical analyses and figure generation.

Brief incubation of G-CSF MPBCs with Fas ligand results in selective reduction of CD3 + T cells while maintaining CD34 + viability and functionality
MPBCs from 25 healthy donors were separately incubated for 2 h with hexameric FasL or with control media. Early apoptosis signal and reduction in the percentage of CD3 + T cells were detected in the FasL-treated samples, while CD34 + percentage and viability were unaffected ( Fig. 1a-d). FasL incubation did not affect the percentage of immature CD34 + CD38 low stem cells, multipotent CD45RA − CD90 − stem cells, or self-renewing CD45RA − CD90 + hematopoietic stem cells [28] (Fig. 1e-g). Furthermore, FasL treatment did not reduce the number of erythroid and myeloid colony-forming units that formed in semi-solid, growth factor-supplemented media (Fig. 1h). These results suggest a selective effect of the FasL-treatment on CD3 + T cells, with preservation of CD34 + progenitor cell viability and clonogenic potential.
Engraftment of HSPCs is dependent on their ability to home to and be retained in marrow niches [29]. The chemokine CXCL12/SDF1 is responsible for the early phases of HSPCs homing and retention in these niches via binding of its cognate receptor, CXCR4. CXCR4-dependent homing and retention of HSPCs also requires integrins such as VLA-4 and LFA-1 [30]. Ex vivo FasL incubation for 2 h did not effect the expression of CXCR4, or of integrins LFA-1 or of VLA-4 on CD34 + HSPCs. In functional migration assays, CD34 + HSPCs migration towards a CXCL12/SDF1 gradient was not effected by ex vivo FasL treatment ( Supplementary Fig. 1a, b).
Engraftment rates of xenotransplanted human leukocytes and hematopoietic progenitor cells (percentage of human CD45 + and CD34 + CD38 low cells, respectfully) derived from either control or FasL-treated-MPBC grafts were similar in marrow aspirates from NSG mice 4 weeks after transplantation (Fig. 1i, j). Furthermore, similar percentage of mature human CD19 + B cells, CD33 + and CD14 + CD16 − myelo-monocytic, as well as CD56 + CD16 − NK cells were observed following control-or FasLtreatment of xenografts (Fig. 1k). A small but statistically significant increase in the number of human CFU was observed in the BM of mice transplanted with CD34 + cells derived from FasL-treated-MPBCs compared to the control treatment group at 4 weeks (Fig. 1l). These results demonstrate that ex vivo FasL treatment does not impair short-term engraftment and differentiation of FasL-treated-MPBCs.

Fas ligand induces apoptosis in specific T cell subsets, reduces expression of activation proteins, and attenuates IFN-γ secretion
GvHD is initiated by alloreactivity generated within specific donor-derived T cell subsets; we examined the differential apoptosis-inducing effects of brief ex vivo FasL incubation on T cell subsets present in MPBC grafts. Rates of FasLinduced, early apoptosis were highest among naïve T stem cell memory (T SCM ), mature helper (T H , CD4 + ) and cytotoxic (T C , CD8 + ) T cell populations (Fig. 2a, b). Differential FasL sensitivity of each T cell sub-population correlated with the intensity of expression of Fas receptor (CD95) on the cells of each specific population (Fig. 2c).
To simulate the allo-activation that occurs following infusion of a stem cell graft, CD3 + cells derived from MPBC or FasL-treated MPBC were exposed to human T-activator CD3/CD28 beads for 24-48 h; the expression of T-cell activation markers was tested. Incubation of MPBCs with FasL before T cell stimulation reduced the number of CD25 + cells (Fig. 3a, b) and significantly reduced both secreted IFN-γ as well as the percentage of cells expressing IFN-γ (Fig. 3c, d). Brief incubation with FasL reduced the percentages of T H 1, CD4 + CXCR3 + cells, T C 1, CD8 + CXCR3 + cells and T H 17, CD4 + CCR6 + cells (Fig. 3e-h). Interestingly, we found no difference in the percentage of either CD4 + CD25 + FoxP3 + or CD8 + CD25 + FoxP3 + T regulatory cell populations following ex vivo treatment with FasL (Fig. 3i).

FasL-treated-MPBCs attenuate acute GvHD and improve survival following xenogeneic transplantation
We studied human MPBC engraftment and the development of GvHD in γ-irradiated NSG mice transplanted with either FasL-treated or control grafts. On day 18 post transplantation, the mean clinical GvHD score of control MPBC-infused mice was 6.1 ± 1.0, manifest as weight loss (Fig. 4a), hunched posture, ruffled fur and reduced locomotion. By contrast, the mean clinical GvHD score of mice infused with FasL-treated-MPBCs was zero (P < 0.001) (Fig. 4b). GvHD progressed in control-MPBC mice, and none survived to 28 days post-transplantation; by contrast, all mice transplanted with FasL-treated-MPBCs survived at the 60-day benchmark (Fig. 4c) GvHD is triggered early after transplantation when donor T cells home to secondary lymphoid organs, where they undergo activation and expansion before migrating into GvHD target organs [6]. We monitored splenic infiltration by human lymphocytes as a surrogate measure of in vivo T cell expansion and lymph node homing after xenotransplantation of human FasL-treated or control MPBCs in irradiated, immunodeficient mice; human hematopoietic cell engraftment in marrow and peripheral blood were monitored at the same timepoints. Spleens harvested 7 and 14 days after infusion of FasL-treated-MPBCs were significantly smaller (Supplementary Fig. 2a) and contained fewer hCD45 + , hCD3 + , hCD19 + , and hCD33 + cells as compared to spleens of mice infused with control MPBC grafts ( Fig. 4d-g, respectively). In contrast, hematopoietic marrow engraftment, as measured by the number of human  CD34 + cells (Fig. 4n) and myeloid precursors in marrow aspirates (Fig. 4k), was similar at day 14 in both control and FasL-treated-MPBC transplanted mice. Additionally, human T and B cell numbers were reduced in the marrow of recipients of FasL-treated-MPBCs as compared to control mice ( Fig. 4i and j) as where hCD45 + and hCD3 + cells in the peripheral blood (Fig. 4l, m). Reduced clinical GvHD scores of FasL-treated grafts were paralleled by lower IFN-γ plasma levels (Fig. 4b, o) and a reduction in histologic signs of GVHDs in the liver parenchyma of these mice ( Supplementary Fig. 2b, c).

Fas ligand induces apoptosis in B cell subsets
Accumulation of T cells in the spleens of control MPBCs recipients suggests the presence of local antigen-presenting cells (APCs). Indeed, we noted a parallel increase in splenic human T and B cells in our xenograft recipients following the administration of control MPBCs, and attenuation of this process in recipients of FasL-treated-MPBCs (Fig. 4f). As B cells serve as surrogate APC's in xenograft models [31,32], B cell subsets were evaluated in FasL-treated-MPBC grafts (Fig. 5a-c). FasL treatment lead to increased apoptosis of transitional, naïve, memory and plasmablastoid B cells and a reduction in their relative percentage within the treated grafts (Fig. 5a, b). CD95 expression was noted in the most affected cell populations (data not shown). Ex vivo FasL treatment selectively reduced the percentage of HLA-DR high B cells (Fig. 5c). Two weeks after transplantation in immune-deficient mice, transitional, naïve and plasmablastoid B cells were reduced in recipients of FasL-treated-MPBCs (Fig. 5d) as compared to recipients of control grafts. HLA-DR high B cells were selectively decreased in spleens of mice transplanted with FasL-treated-MPBCs Due to high donor-dependent variability, one representative experiment out of 3 is presented. Statistical analysis was performed using unpaired, parametric Student's t test *P ≤ 0.05, **P ≤ 0.01, ***P ≤ 0.001, ****P ≤ 0.001. as compared to mice transplanted with control MPBCs (Fig. 5e), mirroring the results of the ex vivo experiments noted above.

FasL treatment maintains GvL activity while preventing GvHD
The presence of T cells in the transplanted graft promotes both engraftment and GvL [33]. To study the effect of FasL treatment on GvL in vivo, we developed a novel model for concurrently testing GvHD and GvL in NSG mice. MV4-11 human leukemic cells were administered intravenously into γ-irradiated NSG mice on day 0 (10 × 10 6 cells/mouse), and either FasL-treated or control MPBCs (3 × 10 6 TNCs/mouse) were infused 4-6 h later. GvHD scores were recorded twice weekly for three weeks and at the timepoint at which the mice were sacrificed; leukemic burden in the marrow, spleen and blood was assessed using antibodies to human CD123 (Fig. 6a, b). As compared to mice infused with sham stem cell grafts (vehicle), leukemic burden was similarly diminished in the spleen, marrow, and blood of mice co-transplanted with either FasL-treated-MPBCs or control MPBCs (P < 0.01 either transplant vs. sham) (Fig. 6c-e), suggesting that xenograft-mediated GvL was not compromised by FasL treatment. As in the previous experiments (see Fig. 4), recipients of FasL-treated-MPBCs continued to be robust, with no weight loss and with lower GvHD clinical scores as compared to mice receiving control MPBCs (Fig. 6f, g). Of note, vehicle infused mice showed worsening clinical scores due to increasing leukemic burden, and not due to GvHD (Fig. 6g). In summary, FasL permits uncoupling of GvHD and GvL effects in this xenograft model.

Discussion
In this study, we have shown that brief incubation of human stem cell grafts with FasL uncouples GvHD and GvL   Representative results of one independent study out of two is presented. Data presented as Mean+SD, n = 3. Statistical analysis was performed using unpaired, parametric t-test; *P ≤ 0.05, **P ≤ 0.01, ***P ≤ 0.001. effects without attenuating the engraftment potential of the cells in irradiated, immune-deficient mice. Attenuation of GvHD by brief incubation of otherwise unmanipulated human stem cell grafts with multimerized FasL occurred even in the absence of prior priming of donor T cells with host antigens. FasL incubation increased apoptosis of IFN-γ producing T cells and of cell populations expressing high levels of CD95, including naïve derived T SCM , mature helper (T H , CD4 + ) and cytotoxic T cells (T C , CD8 + ), memory and effector T cells, and T H 1/T C 1 and T H 17 cells (Fig. 2c). Recent studies suggest that acute GvHD is mediated in part by mature donor T H 1 and T H 17 cells present in the allograft that recognize minor or major histocompatibility disparities between donor and host [34,35]. In addition to fully differentiated T-cell populations, T SCM includes a minor population of self-renewing cells which expands in the host early after transplantation [36][37][38], and the dynamics of T SCM expansion and contraction correlates with GvHD-free survival [39,40]. These cells are at the cusp of the transition between naïve and effector state and may serve as a cellular reservoir for GvHD-inducing cells. It is not surprising, therefore, that partial depletion of these cells from the graft by FasL treatment would ameliorate GvHD. Tandem depletion of less differentiated T SCM cells and mature, IFN-γ secreting T cells may shape the alloreactive post-transplant T cell milieu, reducing the risk of severe GvHD. Recent descriptions of the role of T cell collectivity and the influence of small T cell populations on antigen response suggest that even partial depletion of small but influential T cell subpopulations within the graft's lymphocyte compartment may have substantial effects on post-transplantation alloreactivity. Further elucidation of the rules of T cell social behavior will likely continue to inform the future development of novel GvHD therapeutics [41].
During stem cell engraftment, donor T cells interact with both host and donor APCs; host APCs present host major and minor histocompatibility antigens to donor T cells priming the acute GvHD reaction [6]. T-cell recognition of MHC molecules is species-restricted; human TCRs do not recognize mouse MHC. GvHD in xenogeneic transplant of human MPBCs into immunodeficient NSG mice ensues when HLA proteins on APCs contained in the graft process and present mouse antigens to T cells that are also contained within the graft [31,32]. Indeed, in our model, donorderived T and B cells co-accumulate in the spleen presaging the appearance of GvHD (Fig. 4d). Brief incubation of MPBCs with FasL resulted in increased apoptosis among transitional, naïve, memory and plasmablastoid B cells, leading to a reduction in their percentage within the treated graft (Fig. 5a, b), and, eventually, within the spleen of transplanted mice (Fig. 5d, e). Furthermore, FasL treatment selectively reduced the percentage of putative antigenpresenting HLA-DR high expressing B cells in vitro and in vivo (Fig. 5c, e). Induction of apoptosis and reduction in the number of myeloid APCs that express HLA-DR high was found in vitro as well as in vivo in the spleen of mice transplanted with FasL-treated grafts ( Supplementary  Fig. 3). This FasL-induced decrease in antigen-presenting capacity corelated with lower GVHD scores and mortality.
We did not directly measure markers of inflammation in treated and control mice, but speculate that FasL-mediated changes in graft composition may reduce levels of inflammation and promote short-term HSC engraftment and hematopoietic recovery.
Host APCs play an important role in the balance between the induction of GvHD and the development of tolerance [42]. Phagocytosis of apoptotic cells converts immature dendritic cells (DCs) into tolerogenic DCs, abrogating GvHD [43,44]. Several different techniques can be used to increase the number of apoptotic cells within the graft. Apoptosis can be induced by incubation of the graft with a photoactivated dye (8-Methoxypsoralen) followed by exposure to ultraviolet light, a process which is exploited clinically in extracorporeal photopheresis (ECP) in GvHD patients unresponsive to conventional immunosuppressive drugs [45], and FasL-induced apoptosis of lymphocytes contained the grafts used in these studies may lead to immunomodulation of macrophages and DCs and reduced rates of GvHD [46,47]. Support for this theory is provided by suppression of GvHD by FasL pretreatment of transplanted grafts in a model that uses syngeneic mice [24]. Ex vivo T cell depletion from donor grafts has been used to prevent GvHD for more than 40 years. Iterations of this technique include positive selection of CD34 + hematopoietic stem cell progenitors and immunomagnetic depletion of target T cell populations [7]. Other approaches have been developed to deplete GvHD causing cells in vivo using anti-T-cell antibodies, or the administration of posttransplant cyclophosphamide [48]. Despite the positive effects of GvHD reduction seen after the administration of these modalities, patients often pay a price for T cell depletion, manifest by delayed immune recovery and attenuation of GvL effects. Brief incubation (2 h) of MPBCs with hexameric Fas ligand, offers a novel ex vivo technology to treat GvHD by selectively depleting specific subsets of T cells, together with the ability to attenuate donor and host allo-antigen presentation involved in GvHD. Engraftment in our murine model is robust following transplantation of treated graft, and the graft retains its immune reconstitution and anti-leukemic effects A phase I clinical trial using such graft in adults undergoing stem cell transplant for the treatment of hematological malignancies is currently underway (NCT02828878).
Acknowledgements The authors thank Alla Zozovsky and Anna Zemlyansky for their excellent technical support.
Author contributions HL-B, LS-P, GR, AS, MR, LG, AR, NM, MA and IM performed the experiments and collected the data used in this paper. RB-K, SY, HW, TK, YB, TZ and JS conceived and designed the experiments described in this paper and analyzed the data produced. Amnon Peled contributed data and analytic tools. The paper was written by HL-B, LS-P, AP and JS. All authors read, commented on and approved the manuscript.

Compliance with ethical standards
Conflict of interest HL-B, LS-P, GR, AS, MR, LG, AR, NM, RB-K are employees of Cellect Biotherapeutics, and SY is an employee and shareholder of the same concern. AP serves as paid consultant for Cellect Biotherapeutics. MA, IM, HW, TK, YB, TZ and JS declare no competing financial interests. This technology is covered by patents and pending patents.
Publisher's note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http://creativecommons. org/licenses/by/4.0/.