Skip to main content
European Commission logo print header

Advanced Cell-based Therapies for the treatment of Primary ImmunoDeficiency

Final Report Summary - CELL-PID (Advanced Cell-based Therapies for the treatment of Primary ImmunoDeficiency)

Executive Summary:
Patients suffering from inherited primary immune deficiencies (PIDs) are prone to infections, autoimmunity and cancer. PIDs are caused by mutation in genes (>300) which are essential in the development and/or function of immune cells. Hematopoietic stem and progenitor cells (HSPCs) are populations of cells residing in the BM that support blood cell development, including immune cells. HSPC transplantation from healthy donors is the treatment of choice for severe forms of PID, but in the absence of a compatible donor, its outcome can be limited by delayed or suboptimal immunological reconstitution and complications. Conventional treatment do not offer a definitive treatment modality and carry often a high burden in terms of social costs and quality of life. Advanced therapies medicinal products (ATMPs) based on autologous hematopoietic cells, genetically modified ex vivo through viral vectors represent a promising approach to treat PID and other inherited disorders. The EU-FP7 funded project CELL-PID (Advanced cell-based therapies for the treatment of primary immunodeficiency) aims at clinical application of innovative and safe ATMP to build a healthy immune system in patients affected by PID. The consortium comprises the clinical centers pioneering in Europe in the field of ATMPs, scientists experts in PID and HSPC, along with the current main industrial components involved. Specifically, the project is based on the genetic correction of progenitor cells of the immune system (HSC, lymphoid progenitors) and mature lymphocytes, to be used as immunotherapeutic cells PID patients. CELL PID researchers have developed products that enhance the ability to collect HSPCs from the blood, to expand in vitro, and favour their ability to return back to bone following transplantation. Researchers have also optimised the culture of HSPCs with the addition of specific factors and enrich for T cell progenitors. These products will contribute to immunity building, offering ground breaking opportunities for treatment of PIDs and other conditions in which the immune system is impaired, such as after transplantation. Another piece of the immunity arsenal, is the thymus, the organ supporting T cell differentiation, which can be altered in PIDs. CELL-PID researchers have worked on new strategies to regenerate a normal thymic epithelium in vitro and in vivo to replace the tissue, proving that thymus transplantation is an effective procedure for PID patients who lack this organ. To evaluate potential toxicity of gene therapy, the researchers have developed and validated sensitive yet rigorous assays. Moreover technological tools and equipments have been developed to produce good manufacturing practice-grade of vectors and medicinal products (HSPC) under standardised and industrialised processes. CELL-PID research has produced various viral vectors to stably integrate the specific therapeutic genes into patient cells, targeting 10 different forms of PID. For 3 diseases (severe combined immunodeficienciency (SCID)-X1), Wiskott-Aldrich Syndrome (WAS), and adenosine deaminase (ADA)-deficient SCID clinical trials are advanced while chronic granulomatous disease (CGD) has recently started. Results of these clinical trials show sustained engraftment of gene-corrected cells, restored immune function and general improvement of clinical condition with a positive safety profile. Continuous monitoring of these patients will be important to confirm long-term safety and efficacy. For most of the other PID variants tested in the context of this project, preclinical studies have shown feasibility and proof of safety and efficacy. Clinical trials are planned to be implemented in the next years, also under funding of recently approved Horizon 2020 EU projects.
CELL PID disseminated the knowledge and results at scientific meetings and through >180 peer reviewed publications, organized several workshops and training courses on ATMPs, and has actively fostered student and scientists exchanges between partners. In terms of exploitation and commercialization, main results achieved by CELL PID partners include: i) application for patents; ii) new orphan drug designation for 3 diseases (ADA-SCID, WAS, X-CGD); iii) generating a spin out with biotech and licencing ATMPs for PID through alliance with industry; iv) achieving EU market approval for the first ex vivo stem cell gene therapy for a genetic disease. From a therapeutic perspective, the approaches have now the potential to provide clinical benefit to these otherwise incurable diseases and broaden clinical application of medicinal products able to rebuild and modulate the immune system to acquired immune disorders, allogeneic HSCT and cancer treatment.

Project Context and Objectives:
Primary immune deficiencies (PID) are inherited disorders of the immune system affecting mainly children and in the most severe forms are life-threatening conditions. They are characterized by susceptibility to infections and a tendency to develop autoimmune disorders and cancer. The use of supportive drugs and improvement in clinical care have increased the life expectancy of PID patients. However, conventional treatment do not offer a definitive treatment modality and carry often a high burden in terms of social costs and quality of life. Allogeneic HSC transplantation is the only established curative treatment but in the absence of an HLA-identical donor can be limited by delayed/suboptimal reconstitution and complications. The scientific and technological progress of the past decade has led to the development of advanced therapies medicinal products (ATMPs) based on autologous hematopoietic cells, genetically modified ex vivo through viral vectors. Gene therapy with retrovirally transduced HSC resulted in long-term immune reconstitution and clinical benefit in clinical trials, demonstrating superior survival rates relative to allogeneic HSC transplant donors. The occurrence of genotoxicity in early clinical trials for SCID-X1, Wiskott-Aldrich Syndrome (WAS), and Chronic Granulomatous Disease (CGD) prompted the development of approaches based on self-inactivating (SIN) vectors. In particular, lentiviral vector (LV) offer a unique combination of advantages since they integrate efficiently into a variety of cell types, allowing stable and controlled transgene expression and significantly alleviating the safety concerns associated with previous vector generation.
The main objective of CELL-PID consortium is the clinical application of innovative and safe ATMP to restore a normal functioning immune system in in PID due to genetic defects of the adaptive and/or innate immune system.
This goal has been pursued through basic work on the biology of HSC, lymphoid progenitors and lymphocytes, preclinical studies, and the implementation of clinical trials to produce clinical grade, gene modified, autologous cells. This was made possible through the use of using of state of the art technological platform and the observance of good practice quality regulations at different stages of medicinal product development (GMP, GLP, GCP).
The network comprises clinicians, researchers, biotech and industrial partners pioneering in the field of advanced therapies and aiming at broad clinical application of safe cell-based therapeutic products. Multicenter collaboration was essential throughout the project in order to collect clinical and genetic data on these rare diseases. Moreover, the network benefited from strong links with scientific societies (EBMT, ESGCT), no profit funding agencies and collaboration with national and EU regulatory authorities.
The following outcomes were foreseen at the beginning of the project:
1) The development of new strategies to facilitate homing and engraftment of transplanted stem/progenitor cells, expansion of therapeutically useful and transplantable progenitor cells,
2) The regeneration of thymic activity in PID patients through infusion of ATMP and/or thymic graft
3) The validation of novel advanced therapies using gene modification of HSC, lymphoid precursors or mature lymphocytes through rigorous preclinical efficacy testing and toxicology.
4) The establishment of a sustainable European multicentre platform for the implementation of multicenter preclinical and clinical studies with the newly developed advanced therapies.
5) Transfer of knowledge and technologies to other areas with an unmet medical need.

The activities were subdivided in 13 workpackages (WP) led by different partners. Activities within each WPs were strictly linked to others as protocols, tools, and knowledge were exchanged among WPs and partners. The main activities planned within each WPs are described below:

WP1– HOMING, RETENTION and BM REPOPULATION of HSPC –The bone marrow (BM) is place were hematopoietic stem cells (HSC) normally reside and produce blood cells under physiological conditions. Before an HSC transplant, patients undergo a conditioning regimen to deplete hematopoietic progenitors and promote homing and engraftment of donor HSCs to the BM. Myeloablative conditioning is irreversible if BM function is not rescued by stem cell support while reduced intensity while non-myeloablative conditioning regimens are associated with a lower rate of non-hematological toxicity. WP1 was aimed at pursuing new approaches aimed at enhancing the homing capacity of transplanted HSCs, electively following ex vivo expansion, by optimizing the combinations of immune treatment and/or reduced-intensity conditioning. Moreover, we aimed at increasing the proliferation and retention rates within the BM of transplanted HSCs. Immunodeficient mice engrafted with human HSC from different sources (UCB, BM, mobilized PB) were used a model system.

WP2 – THYMUS RECONSTITUTION –The thymic microenvironment is unique in its ability to promote the development and selection of naïve T cells with a repertoire purged of vital “Self” specificities but prepared to react to injurious “Non-Self”. Essential for this competence are thymic epithelial cells (TEC) which constitute the major component of the thymic stroma. Congenital or acquired thymic stromal defects, in particular TEC-autonomous deficiencies and partial blocks in intrathymic T cell developmental, severely impede and at times complete prohibit the generation of functionally competent and correctly selected effector and regulatory T cells. The objectives of our efforts were therefore (i) to establish a normal thymic microenvironment able to support regular T cell development in pre-clinical models of a select group of primary immunodeficiencies; (ii) to develop novel approaches to commit the progeny of HSC efficiently to a T cell lineage fate; and (iii) to establish improved clinical protocols for human thymus transplantation to correct congenital (and acquired) pathologies marked by defective T cell development and selection due to a cell-autonomous TEC deficiency affecting their cellularity and/or function.

WP3 TOXICOLOGY OF GENE-MODIFIED CELL-BASED ADVANCED MEDICINAL PRODUCTS (ATMPs) –The success of gene therapy for the treatment of PIDs was shown in different clinical trials. However, the dysregulation of proto-oncogenes by insertional mutagenesis is still one of the major risk factors associated with integrating vectors. To avoid the occurrence of proliferative diseases due to insertional mutagenesis, new gene therapy vectors must be tested for their genotoxic potential. The aim of WP3 was to develop new vector designs and to evaluate their genotoxic and phenotoxic potential in different cell-based assays. To achieve this, available in vitro and in vivo test systems should be used, evaluated, optimized and, if possible, humanized. The overall aim of this WP was to develop and further improve biosafety assays for the detection of possible gene therapeutic treatment toxicity and to assist the other Cell-PID partners in analyzing the genotoxic risk of their newly developed gene therapy constructs. Furthermore, we supported the Cell-PID partners in the preclinical documentation of toxicology testing to pave the way for new clinical trials for the treatment of CGD, WAS, Artemis, RAG1 deficiency and X-linked SCID.

WP4 – MANUFACTURING OF VECTORS AND CELLS – The scope of WP4 was to develop, optimize, standardize and eventually industrialise innovative technological tools and processes for the manufacturing of GMP grade medicinal products. The major goals of WP4 were to optimize LV manufacturing and to develop an automated process to enable the magnetic isolation of HSC from blood and BM with their subsequent cultivation and genetic modification with LVs in a functionally closed system. This process was then translated to clinical partners and tested to prove suitability for use in cell manufacturing for clinical trials. Additional activities included: i) optimizatio and scale up manufacturing of vector and HSC; ii) development new methods suitable for the release of ATMPs; iii) production of highly quality purified vector preparations, and iv) manufacturing of vector and cells under GMP for clinical trials.

WP 5 – GENE MODIFIED HEMATOPOIETIC STEM/PROGENITOR CELLS (HSC): THERAPY FOR COMBINED IMMUNODEFICIENCIES –Severe combined immunodeficiency (SCID) is a heterogeneous group of diseases characterised by the absence of T cells. Among them, RAG1, RAG2, Artemis and adenosine deaminase (ADA) deficiencies account for 30% of them. HSCT provides unsatisfactory results when performed with a related HLA partially compatible donor, depending on the clinical status at the time of diagnosis and on the type of SCID form. Gene therapy approaches for SCID-X1, started since March 1999 at Necker Hospital (Paris) and later on at the Great Ormond Street Hospital (London), showed immune reconstitution and sustained clinical benefit but were associated to high frequency of lymphoproliferation. In ADA-SCID patients treated with HSC gene therapy in combination with mild conditioning (San Raffaele Hospital, Milan), adaptive B and T immunological were restored without leukemic events. To increase the efficiency of gene transfer and reduce the risk of insertional mutagenesis, we planned to test the biological efficacy and potential toxicity of SIN LV, encoding for four genes respectively (i.e. RAG1, RAG2, Artemis, ADA). By this approach, the expression of the therapeutic gene was under the transcriptional control of an internal promoter to abrogate the enhancer proximity effect observed in the earlier clinical trials. Studies were performed in vitro using human and murine deficient HSC and in vivo murine models tightly reproducing the clinical phenotype at least for the RAG, Artemis deficiencies. The final goal of this WP was to the implementation of HSC gene therapy based clinical trials for these SCID forms, to restore immune functions in these patients.

WP6 – GENE MODIFIED HEMATOPOIETIC STEM/PROGENITOR CELLS (HSC): THERAPY FOR CHRONIC GRANULOMATOUS DISEASES –CGD is a rare PID characterized by the inability of phagocytes to eliminate ingested pathogens. Affected patients present an elevated susceptibility to bacterial and fungal infections, as well as an excessive inflammatory response that leads to granuloma formation. The most common cause of the disease are mutations in the X-chromosomal CYBB gene, a subunit of the NADPH enzyme complex resulting in deficient antimicrobial activity of phagocytes. Therefore, reconstitution of oxidase activity by gene delivery of an intact gp91phox to autologous HSC is a reasonable approach for the treatment of X-CGD patients lacking a suitable matched stem cell donor. WP6 aimed at the correction of X-CGD using state-of-the-art LVs targeted to myeloid cells. The focus of WP6 focus was on the implementation of transnational preclinical and clinical studies through the following objectives: (i) to develop an innovative approach of gp91phox gene transfer into HSC using SIN LV based on strategies that allow a safe and controlled transgene expression; (ii) to develop strategies to avoid transgene silencing; (iii) to test the toxicity and efficacy of therapeutic LV vector in preclinical models (in vitro and in vivo); (iv) to test the effectiveness of the therapeutic LV in correcting defects in host defense and inflammation and (v) to obtain relevant pre-clinical information for the implementation of a Phase I/II clinical trial.

WP7 – GENE MODIFIED T CELLS: TOWARDS INNOVATIVE THERAPIES FOR IPEX and HLH –WP7 aimed at the generation of autologous gene-modified T cells to correct the T-cell dependent defects in patients with Immunedysregulation, Polyendocrinopathy, Enteropathy, X-linked (IPEX) syndrome and Hemophagocytic lymphohistiocytosis (HLH). The first objectives was to develop gene transfer approaches to correct the immune regulatory defect in IPEX syndrome. To this aim we explored the feasibility and the safety of LV mediated FOXP3 gene transfer in T cells, HSC or lymphoid committed precursors for the induction of functional Treg cells, both in vitro and in pre-clinical animal models. Another important goal was to extend the knowledge on IPEX-like syndromes of unknown origin, associated with immunodeficiencies/immune dysregulation with Treg defects. To this aim we: i) investigated whether the autoimmune manifestations in IPEX-like syndromes are due to quantitative or qualitative defects of Tregs; ii) evaluated whether antigen-specific effector T cells can be converted into Treg cells, in order to exploit the FOXP3 gene transfer technology to modulate immune pathology in a broader spectrum of autoimmune diseases with known auto-antigens. The last objective was to develop a perforin gene transfer system, to correct the immune cytotoxic defect in patients with HLH. To this aim we explored the feasibility and the safety of LV-mediated perforin gene transfer in CD8+ T and NK cells from healthy donors and HLH patients for the induction of functional cytotoxic cells, both in vitro and in pre-clinical models.

WP8 – IMPLEMENTATION OF MULTICENTRE EUROPEAN CLINICAL TRIAS FOR ATMPS in PID –WP8 is focused at implementing European multi-centre clinical trials for PID treatment, and monitoring patients for effective immunological reconstitution and risk of adverse events. At the time of CELL PID initiation, SCID-X1 and WAS with SIN vectors had already completed the preclinical phase, thanks also to the intense collaborative effort developed in the context of a previous FP6 consortium (CONSERT). Multicenter phase I/II clinical trials for these two diseases were thus a central part of the project (see WP9 and WP10). We envisaged that after completing the collaborative efforts of the preclinical efficacy and toxicology phase for other PIDs (see above), at least 3 clinical trials were going to be implemented in the context of the CELL-PID consortium. The selection of new disease was based on information deriving from: 1) basic studies on HSC and thymic reconstitution (WP1, 2); 2) preclinical studies (WP5, 6, 7) within specific disease model; 3) careful toxicology studies using different in vitro and in vivo tests (WP3); 4) outcome of SCID-X1 and WAS clinical trials (WP9, 10). In addition, partners planned to perform also extended follow up studies (immunological and molecular) in PID patients previously treated with retroviral-mediated HSC therapy, to allow comparison of multiple trials using alternative vector and conditioning strategies.

WP9 –PHASE I/II TRIAL OF SIN GAMMARETROVIRAL GENE THERAPY FOR SCID-X1– This WP is dedicated to running multicenter clinical studies for the treatment of SCID-X1. Novel SIN gammaretroviral vectors incorporating the human housekeeping EF1a promoter were developed to overcome safety issues of early clinical trials. This vector retains effectiveness in cell and animal models systems, yet with significantly reduced potential for mutagenesis. The clinical studies were conducted at Necker and UCL, with important contribution from other clinical centers in the US. The study objectives were to evaluate the safety (short-term and long-term), biological activity and clinical efficacy of autologous HSC engineered with a SIN retroviral vector encoding IL2RG. Moreover, we planned to study the in vivo biology of engrafted transduced cells and safety of integration profile.

WP10 –PHASE I/II TRIAL OF SIN LENTIVIRAL GENE THERAPY FOR WISKOTT-ALDRICH SYNDROME– Wiskott-Aldrich Syndrome is caused by mutations in the WAS gene which encodes the cytoskeletal regulator WASP and plays a pivotal role in function of immune cells and platelets. Defective WASP expression has been reported to cause increased bleeding, infections, eczema, autoimmunity, and cancer. WP10 is aimed at performing a multicenter clinical studies with autologous, lentivirally transduced HSC for the treatment of WAS. The main objectives were to assess the safety, biological activity and clinical efficacy of WAS HSC gene therapy in at least 16 patients enrolled in clinical trials conducted in 3 centers (USR/FCSR, Necker, UCL). Since the centers adopted two different conditioning protocol, another important goal was to study the role of different conditioning regimen (myeloablative vs reduced intensity) prior cell transplant.

WP11 – TRAINING AND DISSEMINATION – This WP was dedicated to develop internal personnel training courses in the field of cell therapy standards, and all related societal aspects (ethics, 3R’s). In addition, it will be in charge of disseminating and promote CELL-PID results to the scientific community, regulatory agencies and patient/citizen groups.

WP12 – ETHICS AND REGULATORY ISSUES – The objective of this WP was to monitor and supervise all ethical aspects occurring during the course of the project, particularly related to research in human subjects and children. In addition, we planned to address regulatory issues linked to using ATPMs developed within CELL-PID, in particular to ensure those in full compliance with relevant EU guidance.

WP13 –PROJECT AND INTELLECTUAL PROPERTY MANAGEMENT–WP13 ensured that: i) the work and tasks and reporting are completed on time, within budget and according to high quality standards, ii) a proper management of the dissemination and commercial development of results, iii) an efficient definition of procedures for handling the industrial property rights and the related patent rights as managed by the IPC, iv) creating a database for results and knowledge generated from the research activities and finally v) promoting women participation in CELL-PID.

Project Results:
Optimzing combinations of immune treatment, reduced intensity and/or non myeloablative conditioning as a substitute for total body irradiation.
Three approaches have been evaluated: Mobilizing HSC to evacuate niches (EMC), increasing BM blood vessel permeability (Weizmann) and developing a new immune deficient mouse strain with better human cell engraftment capacity (TUD).
EMC: Engraftment of a limited number of lineage negative hematopoietic cells in the mouse model for X-linked SCID (Il2rg-/-) was studied following mobilization of stem cells in the recipient by G-CSF. G-CSF pretreated recipients displayed full engraftment of WT donor cells and complete correction of the X-linked SCID phenotype. Similarly, G-CSF mobilization conditioning allowed for sufficient engraftment of gene therapy treated X-linked SCID cells to correct the phenotype (Hum Gene Ther. 2014, 25(10):905-14). It is concluded that mobilization of stem cells in the recipient is a valid approach towards a non-toxic conditioning regimen for gene-modified cells in X-linked SCID, facilitated by the selective advantage of normal cells in these mice and promotes B cell recovery which frequently is insufficient without conditioning. The research is continued with combinations of other mobilizing agents as well as studying its applicability in other murine models for inherited disorders with less or no selective advantage of normal cells.
Weizmann: The Blood-BM-Barrier controls homing, retention and mobilization of HSC into and from the BM (Nature. 2016, 532(7599):323-8). Irradiation as a pre-conditioning protocol, in addition to killing host BM dividing cells, reduces permeability of this barrier which allows increased homing rates of the donor infused cells. In order to substitute irradiation we pre-conditioned WT mice with neutralizing anti VE-Cadherin antibodies, to neutralize VE-Cadherin adherent junction molecule expressed by endothelial cells in order to control blood vessel permeability. Mice were transplanted then with fluorescently labeled BM Lineage negative cells from a c-Kit-EGFP donor and donor-type LSK/CD34low HSC homing to the recipient BM were quantified. We found that VE-Cadherin neutralization increased homing by 2-3-fold in comparison to control antibody, suggesting that blood vessel permeability is a good target for substituting irradiation as a pre-conditioning protocol in mice.
TUD: We generated, analyzed and published three new mouse strains suitable for the engraftment of human HSC (Cell Stem Cell, 2014, 15:227-238). In all cases we combined immune-deficiency with a mutation in the Kit receptor referring to a functionally impaired HSC compartment on genetic backgrounds suitable for the acceptance of human HSCs (BALB/c, NOD). Homing and long-term engraftment rates of low numbers of human HSCs are elevated in this mouse strain compared to the best currently available mouse strain (irradiated NSG mice).

To define homing rates, multilineage differentiation and engraftment levels of human HSPC, in immune deficient mice.
Two approaches have been taken: developing new immune deficient mouse strains in which tracking of human homing stem cells and quantifying homing rates is more feasible (TUD), and ex-vivo stimulation of HSC prior to their transplantation in order to improve their homing, engraftment and multilineage differentiation (EMC).
TUD: To facilitate human HSC engraftment, we generated a novel mouse strain, Rag2-/- Il2rg-/- KitWv/Wv mice. These immune deficient mice carry a defective Kit receptor rendering endogenous HSCs functionally impaired. The loss-of-function Kit receptor opens up the stem cell niche across species barriers and allows for robust and sustained engraftment of human HSCs after transfer into adult mice, even without irradiation conditioning. Following stable engraftment in the mouse BM niche, human HSCs give rise to lymphoid, erythroid, and myeloid lineage cells over long periods of time in primary and secondary recipient mice. Therefore, Kit-signaling regulates the replacement of mouse HSCs by a xenogenic blood stem cell graft. In addition, we generated a second mouse model for enhanced engraftment of human HSCs combining the benefits from a mutant Kit receptor resulting in no need for conditioning therapy and sustained high engraftment rates with ease of generation of recipient mice. In this case we used the NSG mouse strain as basis and introduced the Kit-W41 allele that impairs endogenous HSC function but the mice are fertile. Multilineage differentiation in this novel mouse strain NSG-KitW41 was much improved over irradiated NSG recipient mice. More precisely, myeloid lineage are found increased in the BM, blood and spleen suggesting that this model is superior for studying human innate immune responses in a model organism (Cell Stem Cell. 2014;15(2):227-38).
EMC: Evaluated the influence of hypoxia on 7-day serum free hematopoietic expansion cultures stimulated with SCF, TPO, IGF2, Flt3L and Angptl3 (collectively termed STIFA3) for both mouse and human stem cells as tested in syngeneic mouse recipients and NOD/SCID mice, respectively. During 7 days of culture, Lin- Sca-1+ c-kit+ (LSK) cells expanded 80±7-fold under 5% oxygen as opposed to 35±4-fold under normoxic conditions, Progenitor cells and short-term repopulating HSCs expanded 30±5-fold and 33±6-fold in normoxia and 80±8-fold and 75±9-fold, respectively, under hypoxia. Competitive transplantation of the cultured equivalent of 120 or as few as 12 original LSK cells and secondary transplantation of BM cells revealed an approximately 10- to 30-fold expansion of long-term repopulating HSCs, on which, remarkably, hypoxia had no significant effect. In addition, CXCR4 expression on the cells increased some 5- to 7-fold, suggesting that in addition to cell number expansion, homing may be increased as well. Consistent with the mouse studies, NOD/SCID repopulating cells in the CD34+ population of human umbilical cord blood (UCB), stimulated with the equivalent human growth factors under hypoxic conditions, expanded within 7 days approximately 15-fold. These findings provide an essential basis for clinical implementation of ex vivo stem cell expansion for both allogeneic UCB and autologous gene modified stem cell transplantation.
In addition, EMC developed a highly efficient overnight gene transfer (transduction) protocol for HSCs at a low multiplicity of infection (MOI, i.e. the ratio of LV particles to cells) that avoids the currently used pre-stimulation with a cytokines cocktail (Methods Mol Biol. 2014;1185:311-9) and preserves the repopulating capacity of the stem cells by preventing differentiation, since thrombopoietin (TPO) is the only growth factor required during the overnight transduction (manuscript in preparation). By this technology, the efficiency in the use of LV batches will increase 10- tot 50-fold, enabling many more patients to be treated with a single vector batch than is applied in the current clinical trials, making the transduction procedure also due tot the limited growth factor requirements highly cost-effective. In the course of the project, the used culture medium was successfully replaced by Miltenyi’s StemMACS™ medium. At subsequent studies at HU the clinical grade version of this medium was found to be equivalent both for HSC transduction and expansion, which opens the way to GMP-compliant clinical implementation in upcoming gene therapy trials.

To increase levels of SDF-1 in the host BM in order to facilitate the homing of transplanted HSC.
Two compounds were tested for the potential to increase SDF-1 levels in the mouse BM.
WEIZ: The bioactive sphingolipid metabolite Sphingosine 1-Phosphate (S1P) is involved in migration of lymphocytes. The chemotactic effect of S1P on mouse BM progenitors was first documented in vitro. Administration of DOP (an inhibitor of S1P enzymatic degradation) to mice was found to increase SDF-1 expression and production in their BM with no effect in the blood. Physiologic levels of corticosterone were found to regulate mouse BM HSC proliferation and mobilization (Leukemia. 2013; 27,2006–2015). Reducing the levels of Corticosterone by Metyrapone injection in mice increased their SDF-1 in the BM and PB. Higher levels of SKL stem and progenitor cells were found in NOD/SCID treated mice. Metyrapone injections were tested as a pre-conditioning protocol which increased homing of mouse BM mononuclear cells to the BM of non-irradiated recipients.

To manipulate HSPC, including ex vivo expansion, prior to their transplantation in order to increase their homing capacity.
Two approaches have been taken to expand HSC ex vivo: Co-culture with human BM mesenchymal stromal cells (MSC) endowed with the highest potential to support HSC in co-culture, and cytokine stimulation.
HU: We previously used the cytokine combination STIFA5 (SCF, TPO, IGF-BP2, Flt3-L, Angptl-5) to expand human HSCs in presence or absence of MSCs, isolated by plastic adherence (PA-MSC) or prospectively isolated using CD271. The EMC group previously found that STIFA3 (SCF, TPO, IGF2, aFGF, Angptl-3) was optimal for expansion of murine HSCs. We compared both protocols and performed dose optimizing studies. Results demonstrate that STIFA3 and STIFA5 combinations are similar in their ability to support ex vivo expansion of HSCs. Human BM CD271++/CD45- stromal cells were found to contain all CFU-F capacity among other stromal cell subpopulations. Differentiation capacity of CD271++/CD45- cells was similar to plastic adherent MSCs obtained through direct plating of BM-MNCs. CD271++/CD45- cells supported expansion of CD34+ UCB cells. Cultures in absence of feeder layers, showed in general higher fractions of CD34+/CD38- cells after 7 days of culture (65.7%) in comparison to 35.4% CD34+/CD38- cells in co-cultures, whereas CD34+/CD38+ fractions are relatively large after co-culture (59.0%) versus 21.7% in cultures without feeder layers. However, due to the overall fold increase in total cell numbers, the absolute numbers of CD34+/CD38- cells are highest in the co-cultures. Cells expanded for 7 days on CD271 feeder layers, were next differentiated up to 65 days on Op9 and Op9-DL1 feeder layers to induce lymphopoietic differentiation in presence of IL-7 and Flt3 ligand. Cells expanded on CD271+ cell layers displayed differentiation towards early T (CD1a positive cells) and B (CD10 positive cells). To prevent interfering with expression of homing molecules during the process of HSC recovery from HSC/MSC co-cultures, we compared the effects and efficacy of different enzymatic and non-enzymatic cell detachment methods/solutions on the expression of CXCR4 by human CD34+ HSCs and BM MSCs as well as the recovery of CXCR4 expression. An optimal procedure has been calibrated accordingly.
Weizmann: In vitro treatment of mouse BM cells with the cytokines bFGF and PGE2, increased CXCR4 expression, cell motility and long term repopulation capacity via reduction of Reactive Oxygen Species (ROS) in the HSC. BM LT- HSC that highly express EPCR, contain low levels of nitric oxide (NO), rendering reduced cell motility and VLA4-dependent high adhesion, to support their attachment to niche supporting stromal cells in the BM (Gur-Cohen, Nat Med, 2015). EPCR-expressing mHSC homed preferentially to the BM but not to the spleen of transplanted mice, which could be blocked by EPCR neutralization. Homing increase could be facilitated by pre-stimulation with the EPCR-ligand, aPC. Among human CD34+ from various sources the more primitive CD34+CD38- cells contain lower levels of NO and ROS. Further reduction of ROS levels in vitro by the ROS scavenger N-Acetyl Cysteine (NAC) reduced SDF-1-directed migration, and in vivo homing in NSG mice. Similarly, reducing NO production decreased SDF-1-directed migration of CB CD34+CD38- cells. Contrary, treatment with thrombin, enhanced NO generation, upregulated CXCR4 expression on human CB stem and progenitor cells and enhanced their homing to the BM and spleen of transplanted immune deficient mice. Our results reveal that human UCB stem cells functionally express receptors of the coagulation system which regulate on the one hand their BM retention and maintenance and on the other hand their motility and development.

To increase proliferation rates of homing cells
Weizmann: In vivo 7 days treatment with the cytokine bFGF of NOD/SCID mice pre-engrafted with human UCB cells induced human CD34+CD38- HSPC in the murine BM into cell cycle. The same was found in WT mouse settings, where bFGF in vivo treatment expanded the BM HSPC pool. In addition, bFGF in vivo administration induced proliferation of BM Nestin+ MSC, niche cells known to support hematopoiesis. We also found that in vivo treatment with bFGF and PGE2, expand ROSlow HSC in the BM. We recently identified that mouse BM EPCRhigh HSC contain low levels of nitric oxide (NO). We further tested whether limiting NO levels in vivo can increase long-term-HSC repopulation potential. Prolonged in vivo treatment with the drug L-NAME (inhibitor of the NO-producing enzyme in HSC, eNOS) reduced immature LSK progenitor cell egress to the blood and preferentially increased the levels of primitive and quiescent EPCR+ LSK HSC in the BM. Similarly, in vivo treatment of mice with the EPCR-ligand, aPC, reduced mature and immature leukocyte egress to the blood, while preferentially increasing quiescent CD34− LSK and primitive BM EPCR+ LSK cell populations within the BM. This treatment further improved the long-term repopulation potential of HSC collected from the BM of treated mice, assayed in a competitive repopulation assay (Gur-Cohen, Nature Medicine, 2015).
HU: Human BM plasma and cells from endosteal and vascular regions from the BM of G-CSF stimulated (n=10) and unstimulated healthy donors (n=10) were collected for comparison to assess differences in expression of hematopoietic growth factors, cytokines and chemokines. IL-1b, IL-3, IL-8, IL-18, and CCL5 were upregulated while CCL11 and CCL27 were reduced following G-CSF treatment. IL-3, IL-4, CCL5 and TNFa were lower at the endosteal region, and PDGFbb was higher at the vascular region. While no differences in cell surface markers of mesenchymal or HSC, PDGF-Rb expression was significantly higher in MSCs derived from the endosteal niche ve MSCs from the vascular niche. Endosteal MSCs showed differentiation towards osteogenic lineage, whereas MSCs from the vascular region showed enhanced differentiation towards adipogenic lineage. G-CSF treatment of patients increased PDGF-BB secretion which may affect MSC expressing the receptors PDGF-AA and PDGF-BB. In vitro, PDGF-BB dose-dependently suppressed MSC proliferation by dowregulation of PDGF-Receptor expression and induction of apoptosis. Similar effect was obtained by PDGF-R inhibitors or a tyrosine kinase inhibitor which also targets PDGF-R.

To find novel targets for alternative pre-conditioning protocols.
TUD: Long-term HSCs from adult mice that lack cell surface expression of the Kit receptor (KitW/W human erythropoietin [EPO]-transgenic) and control mice (EPO-transgenic and wild type (WT) were sorted based on the Lineage- Sca-1+ CD48- CD150+ CD135- CD34- cell surface phenotype and gene expression analysed by next generation sequencing. After normalization, testing for differential expression was performed using Excel by calculating the fold-change, and accepting a maximum p-value of 0.05. Differentially expressed genes that are unique to Kit-surface null HSCs were obtained by subtracting these genes induced by EPO overexpression compared to WT HSCs. Based on this analysis, the Lin28b-let-7-Hmga2 axis was found transcriptionally depressed in Kit-mutant adult HSCs, confirming the role of that pathway in adult HSC function. Other targets potentially involved in HSC function are: Cyp26b1, Tox, Runx1t1, Angpt1, Ppargc1a, and Hlf.

Partner 1 - USR: The thymic epithelial compartment in the Omenn mouse model has been characterized before and after HSCT and upon gene therapy treatment. We have demonstrated that Omenn mice show the absence of cortico-medullary demarcation, reduction of the cortico-medullary ratio and dramatic loss of medullary thymic epithelial cells. Following haematopoietic engraftment with WT BM cells, we found a complete restoration of the epithelial compartment demonstrating the appearance of cortico-medullay demarcation with expression of Claudin4+ UEA+ cells, and the presence of AIRE+ cells indicating the full maturation of medullary thymic epithelial cells. We have also demonstrated the restoration of T cell differentiation and the generation of CD25+ Foxp3+ cells. In parallel, the analysis of the thymus obtained from Omenn mice engrafted with hematopoietic cells in which by gene replacement therapy functional Rag cDNA was introduced demonstrated a reduction in the proportion of DN cells with a concomitant increase in double positive thymocytes that have ostensibly overcome the maturational block typically present in uncorrected thymocytees from Omenn mice. In parallel, we have evaluated the effect of cytokine administration. To this end, we evaluated Keratinocyte Growth Factor (KGF), demonstrated to have a proliferative effect on thymic epithelial cells (TECs) and leading to a more efficient generation of thymocytes. On this basis, KGF has been proposed in therapy to boost thymic reconstitution in HSC transplantation in young and old mice. We have evaluated the efficacy of KGF in adult and new born Omenn mice. To this end, five weeks old OS mice were injected with three consecutive doses of KGF (5mg/kg) weekly. Thymic reconstitution was evaluated at different time points (7, 14, 21, 28 days). Thymic cellularity and stromal differentiation, cortical thymic epithelial cells and medullary thymic epithelial cells, were analysed in treated mice and controls. Preliminary data indicate a significant increase in thymic cellularity, particularly in absolute counts of all thymic epithelial populations comprising the cortical and the medullary one. In parallel, we have treated newborn OS mice. Similar results have been observed in OS mice receiving treatment at two weeks old, whereas no epithelial improvement was noted in new-borns starting treatment at three days old. An increase, even at not statistical level, in CD4 and CD8 absolute number was noted in the absence of any changes in their immunophenotype. Finally we have evaluated the effect of anti-CD3 mab administration on thymic reconstitution in Rag2 knock-out model and in Omenn mice (Rag2R229Q). We demonstrated the ability to this antibody to induce thymic expansion and cortico-medullary demarcation. These animals, in spite of the inability to induce the autoimmune regulator, displayed a significant amelioration in thymic epithelial compartment and an important reduction of peripheral T-cell activation and tissue infiltration. Furthermore, by injecting a high number of RAG2R229Q progenitors into RAG2-/- animals previously conditioned with anti-CD3ε mAb, we detected autoimmune regulator expression together with the absence of peripheral immunopathology. These observations indicate that improving epithelial thymic function might ameliorate the detrimental behavior of the cell-autonomous RAG defect. Our data have provided important therapeutic proof of concept for future clinical applications of anti-CD3ε mAb treatment in SCID characterized by poor thymus function and autoimmunity.

Partner 2 - INSERM: INSERM has provided proofs of their ability to generate ex vivo T-cell progenitors from UCB derived CD34+ cells by exposing them during few days to the Notch ligand DL-4 and cytokines. This aim was achieved by demonstrating that a seven-day exposure of UCB CD34+ HSPC allowed (i) the generation of high numbers of CD7 expressing T-cell progenitors phenotypically resembling human thymic precursors, (ii) with no T-cell receptor rearrangements detected at this time point, (iii) expressing several genes (Rag1, pTa, Bcl11b and IL7Ra) crucial for T-cell maturation, (iv) with a highly increased T-cell potential in vitro (>200 fold as compared to non cultured HSPC). When transferred into NOD/SCID/γc-/- mice, DL-4 primed T-cell progenitors migrated to the thymus and developed into mature αβ T-cells that subsequently left the thymus and accelerated T-cell reconstitution. We also demonstrated that human T-cells circulating in the periphery of the recipients of DL-4 T-cell progenitors were polyclonal and functional in vitro. The conditions of culture on DL-4 were further improved by the addition of an adhesion factor, change of the medium and the modification of cytokines concentrations. The yield of T-cell progenitors obtained was multiplied by two folds. This optimized culture conditions were then applied to adult HSPC (sorted from mobilized PB of healthy donors). A peak of T-cell precursors was reached after 3 days of culture. They were compared to UCB day-7 derived T-cell precursors in terms of in vitro differentiation potential and molecular profiles. They exhibited the same T-cell potential (with a frequency around 1 out of 10). Furthermore transcriptomic analysis revealed a similar induction of the T-cell differentiation molecular program concomitant to silencing of other hematopoietic programs as UCB counterparts.
According to the original task plan, INSERM partner started in vivo experiments with T-cell precursors generated from adult HSPC. Human DL-4 T-cell precursors generated from HSPC were transplanted into adult Busulfan-conditioned NSG in two independent experiments. Controls were transplanted with HSPC not exposed to DL-4. Analysis of the recipients showed that as soon as 6 weeks post-transplantation, the thymus and the spleen of the recipients of DL-4 progenitors were full of mature T cells in contrast with the « empty » thymus and spleen of controls recipients. These results indicate that, like their neonatal counterparts, ex vivo generated, adult T-cell progenitors are able to give rise to a quick wave of thymopoiesis and thus to accelerate T-cell reconstitution.
A phase I/II clinical trial is planned for 2017 and will used adoptive transfer of in-vitro generated T cell precursors to favour immune reconstitution in allogeneic transplantation (See Section 3).

Partners 3 & 8 - UCL/ OPBG: In total 20 patients (including one intention to treat case dying before transplant) were enrolled during the grant period to receive a human thymus transplant. Sixteen of these patients are currently alive, which represents a survival of 80%. Six of the 11 surviving patients that have reached a sufficient follow-up to be evaluated have developed autoimmunity as a complication following the engraftment of allogeneic thymus tissue and the establishment of a T cell compartment. Autoimmunity mostly affected the thyroid gland and blood components. Pre-existing viral infections were the most common cause of death, a finding that confirms results obtained from the only centre that performs thymus transplants in the US.
Timely access to thymic tissue for transplantation can be limiting given that the graft will need to be thoroughly quality controlled. It is for this reason that attempts were initiated to cryopreserve thymic tissue slices. The results of this effort now show that thymic tissue still retains high viability (80-90%) following crypreservation and thawing and, importantly, can retain function as verified in a mouse model.
Intrinsic thymic stromal defects are only one of several reasons that can cause a partial or complete absence of peripheral T cells. Thymus transplants will not rescue numerical T cell deficiencies that originate from hematopoietic pathologies and, conversely, intrinsic thymic stromal defects cannot be corrected y haematopoietic stem cell transplantation. To assess whether the absence of mature T cells is caused by a thymus stromal defect in a patient in which a precise molecular diagnosis for the observed immunodeficiency has not (yet) been possible but therapeutic correction of the immunodeficiency is urgently warranted, we developed co-cultures of BM cells and Op9 Delta 1 cells, a bone-marrow-derived stromal cell line that ectopically expresses the Notch ligand, Delta-like 1 (Dll1). This approach provides a simple, versatile, and efficient culture system that allows for the commitment, differentiation, and proliferation of T-lineage cells from different sources of stem cells. These co-cultures have now been successfully established as a means to stratify T cell deficient patients and have recently been valuable in identifying a patient with a molecularly undefined stromal defect that will now be able to benefit from a thymus transplantation.
The Italian Italian Network for PID (IPINET) (involving USR/FSCSR and OPBG) completed and published the study on the clinical features and follow up of 223 patients with 22q11.2DS syndrome. The study provides useful guidelines for pediatricians and specialists for early identification of cases that can be confirmed by genetic testing, allowing the provision of appropriate clinical management.

Partner P12-UKBB: The overall objective of our contribution to WP2 was to culture primary mouse TEC whilst maintaining their unique thymopoietic functionality and their capacity to proliferate. We initially tested two different cell culture media, namely CELLnTEC (CNT), which contains hEGF, hFGF-1 and low calcium concentrations, and X-VIVO, which is free of exogenous growth factors, and contains high calcium concentrations. In addition, the media were supplemented with Rho-associated protein kinase inhibitor to minimize dissociation-induced apoptosis of isolated TEC. Both media were initially assessed for their utility to support primary TEC growth and promote the cells’ expansion over an extended period of time (at least 12 days). TECs grown in CELLnTEC displayed higher expansion rate independent of further supplementation with a range of recombinant growth and differentiation factors. These latter had been chosen from experiments in which undifferentiated endodermal epithelia derived from ES cells were stimulated into cells with a TEC phenotype and function. Cells grown in CELLnTEC were then used (with a few exceptions) for all other studies. Importantly, we could establish that by 6 days of culture all cells grown under these conditions expressed the epithelial cell marker EpCAM as well as CD49f and Sca-1, the latter two representing putative progenitor cell markers.
We next tested whether the overexpression of Oct4 in adult TEC would achieve a better expansion of these cells in vitro whilst maintaining their lineage commitment and thymopoietic potential. The rationale for this approach was that Oct4 as a homeodomain transcription factor of the POU family may play a critical role in self-renewal of adult TEC when ectopically expressed. This assumption was based on the acknowledged role of Oct4 in inducing pluripotency and the observation that its ectopic expression in adult mouse somatic tissues results in progenitor cell expansion (but also dysplastic epithelial cell growth). To induce ectopic Oct4 expression we took advantage of a third-generation LVs and successfully transduced TECs. Cells that ectopically expressed Oct4 expanded significantly better in culture when compared to TECs transfected with LV lacking an Oct4 sequence and used as a control. We next tested whether the enforced expression of Oct4 would alter the expression of Foxn1, a transcription factor essential for TEC specification, differentiation, growth and function. Its functional absence is the molecular cause for the “nude” phenotype characterised by thymic aplasia. Oct4-transduced TEC expressed a higher level of Foxn1 transcripts. Experiments are presently under way to test whether genes regulated by the transcription factor Foxn1 are up-regulated as a consequence of its heightened expression in TEC consequent to Oct4 transduction. We next added to the TEC cultures several additional factors that had previously been identified to drive in vitro the development of embryonic stem cell (ESC) to cells with phenotypic and functional features of TEC. Specifically, we tested activators of the canonical WNT signaling pathway, WNT3a as specific ligand and LiCl, which activates beta-catenin as a downstream effector of the Wnt pathway. We also tested cyclopamine (Cy) to inhibit Sonic Hedgehog signalling, and the fibroblast growth factor 8b (FGF8), which have been demonstrated to be critical for the aforementioned in vitro differentiation. Using a combination of LiCl, WNT3a, Cy and FGF8, increased Foxn1 transcripts under these experimental conditions.
TEC are organized in the thymus in a three-dimensional architecture, where they interact with extracellular matrix (ECM) components that serve not only as cell adhesion sites but also provide key signals for cell growth and functionality. To foster in vitro survival of TEC and improve their rate of expansion, several inert and biodegradable scaffolds were used initially under culture conditions in the expectation that these materials mimic the thymus organotypic organization of its extracellular-matrix. Cultured on these scaffolds, adult TEC survived at a improved rate, and expressed higher levels of MHC-II and the cortical marker Ly51 when compared to the conventional 2D condition.
We next cultured TEC isolated from mice that constitutively express GFP (for easy identification ex vivo) and reaggregated them for 24 hours with thymic stromal cell from day 14.5 embryos. Of note, adult TEC alone are unable to form in vitro cell reaggregate whereas fetal thymic stromal cells have maintained this capacity. Co-culture of adult TEC with fetal thymic stroma allows, however, their efficient three-dimensional growth in an organoid. The rationale for these experiments was to evaluate the capacity of cultured adult TEC to survive, maintain their phenotype and express the normal complementation of essential genes required for regular TEC function after transfer and propagation in vivo. We transferred the reaggregates under the kidney capsule of WT mice and analyzed the grafts by immunohistochemistry and the their capacity to generate T cells by flow cytometry six weeks later. We found that GFP+ TEC were present in the graft and localized both in cortical and medullary region. GFP+ TEC in the cortex expressed the cell-typic cytokeratin 8 (K8), while GFP+ TEC in the medulla expressed K14 which is also characteristically expressed in this compartment. Importantly, GFP+ TEC did express FoxN1 and some also stained positively for the autoimmune regulator (Aire) indicating the adoption of features typical for all TEC and a specific subpopulation of TEC in the medulla, respectively.
In conclusion, the objectives of this project to expand in vitro adult thymic epithelial cells with regenerative and thymopoietic potential were in principal met. By way of ectopic expression of Oct4 and the supplementation of additional growth and differentiation factors we have established condition allowing for a robust expansion of adult TEC in vitro that when transferred in vivo regain a physiological shape, express detectable levels of Foxn1 and can achieve Aire expression. The provision of scaffolds that allow the growth of TEC in a three-dimensional fashion have further contributed to enhanced ex vivo growth of TEC that maintain functional and phenotypic features essential to rebuild a thymus.

The in vitro immortalization (IVIM) assay is one of the few opportunities to test the risk of insertional mutagenesis in cell culture. The IVIM test has been accepted by the European, Australian and American regulatory authorities as a part of the preclinical documentation for investigational new drug (IND) applications or clinical trial applications (CTA), respectively. In this assay, murine HSC are transduced with integrating vectors and exhibit a transformed phenotype in case insertional mutagenesis has activated proto-oncogenes or inactivated tumor suppressor genes in the vicinity of the proviral integrations. We used the IVIM assay to evaluate many different new vector types and architectures. The new vectors were either provided by the different Cell-PID partners or developed in our lab. We performed the assay for lentiviral constructs to treat IL-12, Artemis or RAG1 deficiency, gene transfer vectors to treat β-thalassemia (with or without insulator elements and for different transgenes, including beta-globin or BCL11A), LVs for HIV1 gene therapy and new constructs for GP91, P47 or IL2RG delivery. We also tested alpharetroviral, gammaretroviral and LVs with a battery of different promoters (SFFV, Cbx3-EFS, Cbx3-MND, MND, PGK, UCOE, MFG, LCR-assisted β-Globin promoter) and chromatin insulator elements (cHS4, CTCF).
Despite the broad acceptance of the IVIM assay, there are technical limitations that may diminish the predictive or translational character of the results. The culture conditions clearly favor a myeloid differentiation phenotype. The insertional mutants mostly have integrations in or near MECOM, a clinically relevant proto-oncogene and risk factor for integrating retroviral vectors. However, MECOM activation is not the most prominent clinical severe adverse event in gene therapy trials. Integrations affecting the expression of the transcription factor LMO2 have been proven to be much more severe and relevant, especially when the lymphoid branch of the hematopoietic system is to be treated. Unfortunately, the IVIM assay cannot not be used to screen for this event. Hence, we developed a new assay principle called the surrogate assay for genotoxicity assessment (SAGA) to analyze oncogenic signatures of insertional mutants on the molecular level. We demonstrated that the transformation phenotype of our mutagenic positive control vectors was strongly associated with a specific oncogenic signature. We are in the process of validating and optimizing the new assay, which in principle can also measure the deregulation of LMO2 in dominant clones.
Partner GSH developed a new LV to treat CGD. The construct G1XCGD can be used to transfer a functional copy of the GP91 protein to HSC. First generation LTR-driven gammaretroviral constructs elicited vector induced myeloproliferative disease phenotypes. Integrations in or near the transcription factor MECOM resulted in clonal dominance of insertional mutants, which exhibited chromosomal instability and concurrent transgene silencing. The new LV is devoid of viral enhancer sequences in the LTR (SIN-design) and uses a chimeric promoter to control transgene expression, with low expression activity in HSCs, but a higher expression in myeloid effector cells (see WP6). This design should prevent activation of proto-oncogenes and possible phenotoxicity of the transgene expression in stem cells. MHH used the IVIM assay to test for the mutagenic potential of the new G1XCGD vector and to screen for possible toxicity of the GMP-grade vector supernatant. Our results support the conclusion of a significantly lower transforming capacity of the improved vector and normal viability of the modified HSC.
A major task of WP3 was to evaluate whether mouse transplantation studies with subsequent high throughput insertion site analysis would successfully generate robust and reliable information on the genotoxic potential of new vectors. At present, there are no general guidelines or even a mandatory requirement to perform insertion site studies within a preclinical WP. However, regulatory agencies very frequently ask for these types of analyses before IND application or CTA approval for a new gene therapeutic vector. Moreover, there is a clear need to perform preclinical safety tests under defined parameters, which comply with good laboratory practice (GLP) guidelines. Furthermore, WP3 investigated whether the use of human HSC in a xenotransplantation model is more informative than conventional CD45.1/CD45.2 congenic mouse transplantation studies.
We started our efforts under normal laboratory standards by applying G1XCGD modified murine HSC in the congenic mouse model. We transplanted 47 animals with either G1XCGD or mutagenic control vectors. All mice receiving G1XCGD transduced cells survived with normal health parameters and no signs of vector induced clonal dominance, further substantiating the results obtained by the IVIM assay. The resulting 53 page report was used to complete the preclinical CTA documentation for the vector. The respective clinical trial is now open and recruiting patients.
Similar safety results for other vectors were obtained when the murine transplantation experiments were conducted under GLP-conditions together with the Fraunhofer Institute for Toxicology and Experimental Medicine. Our results showed the feasibility to follow GMP-guidelines during preclinical mouse transplantation studies, but also highlighted challenges associated with the bioinformatic interpretation of insertion sites analyses. We tested several techniques (LM-PCR, LAM-PCR, refree-LAM-PCR, non-restrictive LAM-PCR) to detect and quantify vector integrations in the genomes of transplanted animals. The insertion site retrieval is dependent on the choice of restriction enzymes and nested PCR amplifications, often leading to a significant quantification bias (Brugman et al. 2013). In our hands, linear amplification mediated (LAM)-PCR was the most reliable method for integration analysis, despite the technical limitations associated with the technique. The PCR bias cannot be ignored in the later assessment of the clonal contribution relative to the total amount of engrafted cells. To better understand the integration preferences of different vector genera, the impact of the cytokine conditions on the clonality of the graft and the risk of insertional mutagenesis in gene therapy protocols we invested many resources in the analysis of different lentiviral, gamma- or alpharetroviral insertion site studies (Maetzig et al. 2011; Dahl et al. 2015; Jaako et al. 2014). Even though LVs are considered much safer than their gammaretroviral counterparts, we could demonstrate the functional consequences of vector induced haploinsufficiency of Ebf1 in mice (Heckl et al. 2012). We also successfully detected and analyzed developing clonal dominance in humanized mice (Haemmerle et al. 2014). However, the number of vector modified (human) stem cells in the recipient mice had a substantial impact on the quality of information gained from the insertional profiling in xenotransplanted mice. This was most obvious when human HSC were treated with gene therapy and transplanted into immunodeficient NSG mice. Chimerisms below a certain limit heavily impeded a robust interpretation of the insertion site data (Phaltane et al. 2014).
Flow cytometric methods or qPCR can be used to assess the level of gene marking. We tested SYBR Green as well as Taqman based protocols to measure the vector copy number (VCN) in transduced cells. We served as one reference lab in an inter-laboratory approach for VCN evaluation. Clinical samples from a major clinical study were also assessed in part by MHH (Braun et al. 2014). Our efforts, findings and suggestions regarding the safety of lenti- or retroviral gene therapy were published in two review articles (Rothe et al. 2013; Rothe et al. 2014).

The work conducted in WP4 can be described as two main interconnected sub-projects. The first was related to the development of technological tools to validate the manufacturing of GMP-grade LVs. The second to the development of novel transduction reagents and instruments to industrialize CD34+ HSPCs transduction process with highly purified LVs.
Main achievements obtained for the manufacturing of GMP-grade LVs

1. A pre-GMP pCCL-SIN-GFP vector batch encoding the eGFP marker gene was manufactured and distributed to 8 partners (P1, P2, P3, P4, P7, P8, P11; P15) who participate in the “Stem cell transduction protocols” study. Each partner received 2 ml-vial × 5 containing 1 ml of the DSP 04 purified vector. This study allowed the comparison of different cell culture and transduction procedures in place in the different participating centers to identify the optimal conditions for transduction in clinical batch sizes. This standard protocol was eventually adapted for transduction automation by Miltenyi CliniMACS Prodigy device.
2. Optimisation of sterilizing filtration step in GMP LV vector lot manufacturing to improve safety of the final product. A second filtration step enhances the overall microbial safety of the purified vector because final filtration is performed on a very low bioburden upstream material. Studies focused on the evaluation of possible losses associated with the 2nd filtration step downstream the original unique filtration step.
3. Optimisation of the method for determining the possible presence of Replication-Competent Lentivirus (RCL) in LV lots for clinical application to reduce timing and test costs. This study was made necessary for recent upgrade of process scale and yield and to improve operator safety. The new qualified assay consists in the analysis of 300 ml of initial not-purified vector preparation, instead of the 5% of final purified and concentrated product. This modification was introduced in face of the fact that RCL could be generated only during the vector cultural production phase and therefore it is not necessary to test the precious final purified and concentrated product. The major achievements of the study was the melioration of the infection procedure and the reference HIV-1 R8.71 positive control was produced without VSV-G-pseudotype to improve operator safety.

Main achievements obtained for the industrialization of CD34+ HSPCs transduction process. A closed and highly automated manufacturing procedure would lead to large improvements in product performance of ATMPs, moving cellular and gene therapies from their current translational setting into routine clinical use. We have recently developed a functionally closed and fully automated cell processing device, the CliniMACS Prodigy®, (Figure1) which enables complex cellular products to be manufactured.
The CliniMACS Prodigy maintains a closed system by combining single-use disposable tubing sets equipped with multiple input lines with sterile filters or tubing connections for use with sterile docking devices. Output lines offer in-process control if needed. This closed system meets the requirements of GMP-grade processing of almost any kind of cellular product. It has been developed to fully automate and standardize the manufacturing process of cellular therapeutic agents. The instrument can perform sample loading, cell washing, density-based cell separations including erythrocyte reduction and plasma harvesting, magnetic cell separation, cell culture, and final product formulation.

In this functionally closed system, an integrated workflow for genetic modification of CD34+ cells has been developed. In this transduction process, stem or progenitor cells of interest are automatically washed, magnetically labeled with CliniMACS® CD34 Reagent and enriched using the LP-34 process and the CliniMACS Prodigy Tubing Set TS 310. Following the magnetic enrichment step, the cells are eluted into a cell product bag and this bag is transferred to a second tubing set, TS 730, via sterile docking for cultivation and genetic manipulation. The prototype process developed during the Cell-PID project (Figure 2) allows variables such as cultivation volume and duration, LV volume applied, number of rounds of transduction to be adjusted as required.

In addition to the activities performed on automation, the process of stem cell transduction was analysed in detail. In small-scale experiments the optimal cell culture conditions were achieved to enable an efficient stem cell transduction. As a result, new research and GMP grade stem cell media were developed which demonstrated improved stem cell transduction performance (up to 2-fold increase in numbers of genetically modified cells).
Using the optimal conditions, automated cell preparation, cell labeling and magnetic CD34+ cell separation was performed from both mobilized and non-mobilized apheresis samples as well as BM aspirates. Small scale automated transduction experiments (10e5-10e7 purified CD34+ cells) incorporating a single transduction cycle, resulted in high transduction efficiencies (up to 90%). A scale-up of the procedure was performed using larger numbers of cells (>10e7) enriched from mobilized apheresis material. CD34+ cells were transduced with a clinical-grade LV using a two hit protocol already described in a clinical setting and used as standard for a control experiment run in parallel (Scaramuzza et al. 2013, Mol. Ther), wit gene marking efficiencies in similar range.
An extensive training and technology transfer program to partners at UCL and MolMed/USR was performed, allowing the complete prototype SCT protocol to be established and tested on the CliniMACS Prodigy instrument in comparison to the partner’s established GMP separation and transduction protocols. Comparable performance was observed, both in terms of magnetic separation (compared to historical data sets) and genetic modification efficiencies. Moreover, gene modified cells maintained full engraftment potential as demonstrated by biodistribution studies in NSG (NOD scid gamma) mice. A number of these analyses are still ongoing.
These preliminary data demonstrate that automated cell processing and genetic manipulation of haematopoetic progenitor cells can be efficiently performed in a closed system. However, the usability of the process still needs to be improved to allow routine application in the clinic and this is the focus of further collaborative efforts.

LV gene therapy for RAG1-deficiency. Recombination Activating Gene 1 (RAG1) is involved in activation of immunoglobulin VDJ recombination. RAG1 deficient patient lack mature T and B cells, leading to recurrent infections and a high mortality rate. To correct this deficiency, a LV based gene therapy has been generated in order to express appropriate levels of RAG1 (van Til NP et al., JACI 2014). Briefly, constructs containing the viral spleen-focus-forming virus (SF), ubiquitous promoters, or cell type-restricted promoters driving sequence-optimized RAG1 were compared for efficacy and safety in sublethally preconditioned Rag1-/- mice undergoing transplantation with transduced BM progenitors. PB CD3+ T-cell reconstitution was achieved with SF, ubiquitous promoters, and cell type-restricted promoters but 3- to 18-fold lower than that seen in wild-type mice, and with a compromised CD4+/CD8+ ratio. Mitogen-mediated T-cell responses and T cell–dependent and T cell–independent B-cell responses were not restored, and T-cell receptor patterns were skewed. Reconstitution of mature PB B cells was approximately 20-fold less for the SF vector than in WT mice and often not detectable with the other promoters, and plasma immunoglobulin levels were abnormal. Gene therapy–treated mice had rashes with cellular tissue infiltrates, activated PB CD44+CD69+ T cells, high plasma IgE levels, antibodies against double-stranded DNA, and increased B cell–activating factor levels. Only rather high SF vector copy numbers could boost T- and B-cell reconstitution, but mRNA expression levels during T- and B-cell progenitor stages consistently remained less than wild-type levels. These results underline that further development is required for improved expression to successfully treat patients with RAG1-induced SCID while maintaining low vector copy numbers and minimizing potential risks, including autoimmune reactions resembling Omenn syndrome.
A new LV generated by the Frank Staal team (Pike-Overzet et al., Leukemia, 2011) encodes the codon optimized Rag1 cDNA under the control of ubiquitous chromatin opening element (UCOE). BM CD34+ progenitors cells isolated from RAG1-deficient patients were transduced with this new construct and their capacity to differentiate along the T lymphoid lineage has been tested in vitro on OP96DL1 stroma. In most cases the corrected CD34+ cells did not produce CD4+ or CD8+ T cells under appropriate culture conditions even after prolonged culture (up to 9 weeks), whereas T cells were detected in the UCB CD34+ positive control from 5 weeks of culture. However, some early T-cell receptor rearrangements occurred in the transduced condition as compared to the untransduced patient’s cells condition. All these data suggest that RAG1 expression by this LV is expressed at a too low level and highlight the importance of high RAG1 expression for full VDJ recombination and T cell differentiation. Quantitative PCR showed that RAG1 transgene level was low in all the transduced cells and sustained the T and B cell differentiation only in patient’s cells with very low RAG1 endogenous mRNA. These data suggest that non-functional endogenous RAG1 expression competes with RAG1 transgene expression and prevents restoration of T/B cell differentiation.
These observations indicate that the SIN LV approach for RAG1 deficiency has to be improved before envisaging any clinical implementation. While the required high expression levels might also jeopardize the safety of the approach. Given these considerations, correction of the endogenous deficient RAG1 gene by a gene editing approach should be also considered.

LV gene therapy for Rag2 KO (complete deficiency) and RAG2 KI (Omenn) mice models
RAG2 protein dimers with RAG1 and both play an essential role in VDJ recombination and as a consequence RAG2-deficient patient lack mature T and B cells. We tested the efficacy of SIN lentiviral Rag2 gene transfer in both the RAG2 deficient mouse model and in the Rag2R229Q mice known as a very good model of human Omenn Syndrome (OS).
We developed LVs with the SF promoter driving codon-optimized human RAG2 (RAG2co), which improved phenotype amelioration compared to native RAG2 in Rag2−/− mice. With the RAG2co therapeutic transgene, T-cell receptor (TCR) and immunoglobulin repertoire, T cell mitogen responses, plasma immunoglobulin levels and T cell dependent and independent specific antibody responses were restored. However, the thymus double positive T cell population remained subnormal, possibly due to the SF virus derived element being sensitive to methylation/silencing in the thymus, which was prevented by replacing the SF promoter by the previously reported silencing resistant element (UCOE), and also improved B cell reconstitution to eventually near normal levels. Weak cellular promoters were effective in T cell reconstitution, but deficient in B cell reconstitution. We conclude that immune functions are corrected in Rag2−/− mice by genetic modification of stem cells using the UCOE driven codon-optimized RAG2, providing a valid optional vector for clinical implementation, which is currently prepared at Hacettepe University.
The same vector for the human RAG2 cDNA (UCOE new-RAG2co) has been used to correct the phenotype in the mouse model for Omenn Syndrome. Lineage marker-depleted BM (Lin-) cells were purified from Rag2R229Q mice. Lin- cells transduced with the vector were transferred into sub-lethally irradiated Rag2R229Q recipient mice. The analysis of peripheral immune cell compartments showed a progressive reconstitution of T and B cells associated with a decrease in myeloid cell compartment indicating a normalization of the peripheral immune cell distribution. Spleens of gene corrected mice have a high frequency of B cells in the marginal zone and polyclonal T cell repertoire. Importantly, in the thymus of gene corrected mice, the proportion of double negative cells decreased whereas the double positive thymocyte pool increased assessing that cells had bypassed the maturation blockage. Furthermore, a clear visible cortico-medullary demarcation was present suggesting a restoration of the thymus architecture. In the periphery, T cells were present and functional. Altogether these data showed that LV-mediated gene therapy ameliorates the disease phenotype and represents a therapeutic option for patients with Omenn Syndrome.

LV Gene therapy for Artemis deficiency. Artemis has an essential function in VDJ recombination during T and B cell development. Mice and human data strongly suggest that gene therapy could be considered as an alternative treatment for restoring lymphocyte differentiation and function in Artemis-deficient patients (Lagresle-Peyrou et al., 2006). To test this hypothesis, we have developed within the Cell-PID consortium a LV expressing Artemis transgene that can be used in a clinical protocol. Artemis patient’s CD34+ cells transduced with this construct were able to overcome the blockade in T and B cells differentiation arrest. Furthermore, an extensive in vivo preclinical biosafety study was achieved in a murine model of Artemis deficiency. A GMP clinical batch will be available at the end of 2016, which will be tested on CD34+ Artemis deficient cells in order to confirm the ability of corrected cells to restore T and B cell differentiation both in vitro and in vivo in NSG mice. A new phase I/II gene therapy clinical trial for Artemis-deficient patients will be set up in 2017 thanks to the SCIDNET consortium (FP7 program).

LV gene therapy for ADA-deficiency. We generated a self-inactivating LV with a codon-optimized human cADA gene under the control of the short form elongation factor-1α promoter (LV EFS ADA). In ADA(-/-) mice, LV EFS ADA displayed high-efficiency gene transfer and sufficient ADA expression to rescue ADA(-/-) mice from their lethal phenotype with good thymic and peripheral T- and B-cell reconstitution. Human ADA-deficient CD34(+) cells transduced with 1-5 × 10(7) TU/ml had 1-3 vector copies/cell and expressed 1-2x of normal endogenous levels of ADA, as assayed in vitro and by transplantation into immune-deficient mice. Importantly, in vitro immortalization assays demonstrated that LV EFS ADA had significantly less transformation potential compared to gRV vectors, and vector integration-site analysis by nrLAM-PCR of transduced human cells grown in immune-deficient mice showed no evidence of clonal skewing. These data demonstrated that the LV EFS ADA vector can effectively transfer the human ADA cDNA and promote immune and metabolic recovery, while reducing the potential for vector-mediated insertional mutagenesis. A clinical protocol based on the use of SIN lentivirus expressing the therapeutic enzyme was started at UCL and UCLA (see WP8).

Testing of regulated vectors for XCGD. Regulated LVs for the gene therapy of X-linked CGD were constructed and tested. Two different approaches were followed: At UCL and GSH tissue specific promoters were used to target gp91phox expression to myeloid cells. At USR and OPBG, self-inactivating LVs gp91phox containing either a pgk or a myeloid-specific promoter (MSP) and in addition 2 or 4x miRNA target sequences for miR126 (dual-regulated vectors) were constructed. Both approaches detarget transgene expression from HSC, restricting expression to myeloid cells during cell differentiation. These vectors were successfully tested for their capacity to reconstitute superoxide production in XCGD-deficient cells in vitro and in vivo. Comparative functional analysis of the vectors developed by partners UCL and GSH (pCCL-ChimGp91PRE4 = ChimGP91) and by partners USR and OPBG (PGK.gp91.126T) showed equal efficiency in superoxide production at equal vector copy numbers. After transduction of XCGD PLB985 cells with these vectors, gp91phox positive cells were sorted and tested in a quantitative cytochrome C assay for superoxide production. At vector copy numbers close to 1, both vectors generated superoxide at levels expected to be of clinical benefit for XCGD patients. The ChimGP91 vector was extensively studied in vivo after transplantation of XCGD mice providing restricted expression of gp91phox to granulocytes and monocytes with background levels in B and T cells. Similarly Gp91phox expression was not detected in murine primitive progenitor cells. Among the several LVs constructed by USR and OPBG, one vector containing the artificial, myeloid restricted promoter and 2 copies of the miRNA target sequence for miRNA 126 (MSP.gp91_2x126T) showed targeted restricted expression in myeloid cells with no detectable expression in primitive progenitor cells and reconstitution of superoxide production in human X-CGD patients equivalent to that found in healthy cells. MOLMED produced and purified pre-GMP CGD LV vectors (see WP4), resulting in LV stocks with high titers (> 2.3E+09 TU/ml) and high purity.

Avoiding vector silencing. As silencing of gene expression was seen in clinical gene therapy trials, silencing-resistant LVs containing myeloid specific promoters in combination with an enhancer-less ubiquitous chromatin opening element (UCOE) were constructed and tested for specificity of gene expression and genotoxic effects. We generated smaller version of the original A2UCOE which maintained protection against vector methylation and myeloid specificity when joined to a myeloid promoter in vitro and in vivo.

Efficacy of LV gene therapy in controlling inflammation in animal models of XCGD. Preclinical studies on correction of defects in host defense and inflammation were successfully conducted. We tested the ChimGP91 vector in an inflammation model, whereby X-CGD mice are injected subcutaneously with sterile air into the back. Animals transplanted with ChimGP91-transduced cells resolved the inflammatory insult as efficient as WT animals. Besides the air-pouch model, we developed a bacterial killing assay to directly prove microbicidal activity in gene corrected neutrophils. Using this assay we successfully demonstrated bacterial killing by gp91phox expressing human granulocytes derived from BM CD34+ cells of a CGD patient after in vitro differentiation. USR set up a model of acute infection closely mimicking CGD patient’s airway infection by an intratracheal injection in X-CGD mice of a methicillin-sensitive reference strain (MSSA) of S.aureus. GT with HSCs transduced with regulated LVs encoding CYBB gene restored the functional activity of NADPH oxidase complex and rescued all mice from mortality due to S.aureus as compared to X-CGD or mock-transduced X-CGD mice. Neutrophil infiltrate, bacterial burden and residual lung damage were similar to WT mice. When challenged with bacteria, X-CGD untreated mice showed an exaggerated production of pro-inflammatory cytokines and chemokine, while in GT treated mice inflammation was controlled to levels that were comparable to WT mice.

Studying the effects of chronic inflammation on hematopoiesis in XCGD. We studied the effect of chronic inflammation on the hematopoietic compartment in patients and mice with X-CGD. We found a dysregulated hematopoiesis characterized by increased numbers of hematopoietic progenitors (HPCs) at the expense of repopulating HSCs. In X-CGD patients there was a clear reduction in the proportion of HSCs in BM and PB and they were also more rapidly exhausted after in vitro culture. In X-CGD mice, increased cycling of HSCs, expansion of HPCs and impaired long-term engraftment capacity were found to be associated with high concentrations of proinflammatory cytokines, including interleukin 1 beta (IL-1β). Treatment of WT mice with IL-1β induced enhanced cell cycle entry of HSCs, expansion of HPCs and defects in long-term engraftment, mimicking the effects observed in X-CGD mice. Inhibition of IL-1β signaling in X-CGD mice reduced HPC numbers and partially reversed the engraftment defects. Thus, persistent chronic inflammation in CGD seems to be associated with hematopoietic proliferative stress leading to a decrease in the functional activity of HSCs. Therefore targeted control of inflammation in CGD may be important for a healthy BM compartment and for the development of successful autologous cell therapy approaches.

Preclinical studies: toxicology and biodistribution. Using an in vitro immortalization assay we showed a strong reduction in the mutagenic potential for the ChimGP91 vector when compared to the first generation retroviral vectors. In comparison to the SF91 vector, a construct that shares the LTR and other regulatory elements of the previously used SF71gp91phox vector, the ChimGP91 vector showed a strong reduction in immortalization potential reflecting a low potential activation of proto-oncogenes and immortalization of BM cells. We conducted extensive efficacy testing in murine and human X-CGD cells. For this we used pre-clinical lots of the ChimGP91 vector called G1XCGD. The G1XCGD and MSP.gp91_2x126T vectors were evaluated for biosafety and efficacy in patient derived XCGD-CD34+ cells.
Biosafety and superoxide production in gene modified cells were compared to similar parameters in untransduced XCGD cells and in CD34+ cells obtained from healthy donors in vitro and in vivo in NSG-mice. After transplantation into NSG-mice engraftment levels of transduced human cells were not significantly different from the non-transduced XCGD- and wt-CD34+ control groups. Most importantly about 30-60% of transduced human myeloid cells expressed gp91phox. Furthermore reconstitution of ROS production was observed in neutrophils obtained from BM of transplanted animals and also after in vitro differentiation of hCD34+-cells purified from the BM of transplanted mice. There was no change in hematopoietic lineage distribution or myeloid differentiation properties observed in transduced versus control groups. We also performed in vivo genotoxicity assays for the G1XCGD vector. Mice transplanted with G1XCGD transduced Lin- cells were monitored for body weight, blood parameters and chimerism in regular intervals before being subject to a complete histopathology and 454-clonality analysis after 26 weeks. We found no adverse events after gene therapy with G1XCGD for any of the mice in these studies.
In mice treated with X-CGD Lin- cells transduced with the PGK or MSP regulated vectors, studies in busulfan-preconditioned X-CGD mice injected with transduced X-CGD Lin- cells showed reconstitution of gp91phox expression and NADPH oxidase activity. Hematopoietic lineage distribution in PB, BM, spleen and thymus in transplanted animals did not differ from the lineage distribution observed in animals transplanted with non-modified cells. Studies in NSG mice transplanted with human CD34+ cells transduced with these vectors show effective multilineage engraftment of LV transduced cells.
In conclusion, the LVs G1XCGD and MSP.gp91_2x126T showed a highly improved safety and efficacy profile compared to LTR-driven gammaretroviral vectors, supporting their application in clinical gene therapy trials (see WP8).

Validation of GMP gene transfer protocol. As part of the validation of the CD34+ cell transduction process for the G1XCGD gene therapy trial, INSERM evaluated the impact of CD34+ cell concentration (1x106CD34+/mL or 3x106/mL) on the transduction efficacy. They observed a significant difference in vector copy number (VCN) measured in liquid culture between the two conditions. According to these data, the 3x106/ ml concentration seems to be the optimal concentration to reach for CD34+ cell transduction in the context of the G1XCGD protocol. All these pre-clinical experiments have allowed the Necker staff to submit to the French authorities (ANSM) a clinical trial for 5 patients, which was recently approved.
Development of LV vectors for p47 deficiency. In addition to gp91phox-containing vectors, LVs targeting p47 deficiency, the second most common genetic form of CGD caused by mutations in the NCF-1 gene, were constructed. UCL developed a derivative of the ChimGP91 vector containing a codon optimized version of p47phox. This vector was able to restore high levels of p47phox expression and NADPH oxidase function in monocytes-derived macrophages from p47 patients and in granulocytes obtained by in vitro differentiating murine p47-/- lineage negative cells. Similarly, UZH developed a p47phox LV containing the human micro-RNA 223 promoter, which is known to be strongly upregulated during myeloid differentiation. The miR223 promoter restricted the expression of p47phox to granulocytes and macrophages with only marginal expression in lymphocytes or HSPC. Furthermore, gene correction of primary p47phox patient derived CD34+ cells followed by ex vivo differentiation to neutrophils resulted in restoration of E.coli killing activity by miR223 mediated p47phox expression. These results indicate that the miR223 promoter as internal promoter within SIN LV gene therapy vectors is able to efficiently correct the CGD phenotype with very restricted activity in hematopoietic progenitors limiting the potential of insertional oncogenesis and development of clonal dominance.
IPEX and IPEX-like patients’ characterization and sample collection. Patients with infantile onset diarrhea (after exclusion of infectious causes and not responsive to gluten-free diet), alone or in association with T1D and one or more of the following: skin disease, autoimmune thyroiditis, cytopenias, glomerulopathy/interstitial nephritis, autoimmune hepatitis, polyarthritis, alopecia, increased serum IgE levels, or eosinophilia were included in the present study. Male patients meeting IPEX syndrome criteria underwent FOXP3 genetic analysis. If the presence of FOXP3 mutations was confirmed patients were included in the study as IPEX if the mutation was absent, patients were included as IPEX like. We performed phenotypic and functional characterization of tTreg cells in the PB of the enrolled patients. We showed that quantification of tTreg cells by TSDR/TLSDR combined assay allows the identification of a subset of IPEX-like patients with quantitative deficiency of tTreg cells (Barzaghi et al, J Autoimmunity 2012). Nonetheless, we cannot exclude that in some patients the underlying defect may primarily affect compartments other than tTreg cells.

Furthermore, in order to identify a reliable biomarker to easily distinguish IPEX from IPEX-like patients we screened both groups and control patients with enteropathy for the presence or absence of anti-harmonin and anti-villin autoantibodies (HAA and VAA). Our findings indicate that HAA and VAA are highly sensitive and specific markers for the diagnosis of IPEX syndrome and for the distinction of IPEX and IPEX-like diseases (Lampasona et al., PlOS ONE 2013). In summary, we identified HAA and VAA as major hallmark of IPEX syndrome in the presence of FOXP3 mutations, whereas quantitative Treg cell defect and the absence of HAA and VAA as major features of IPEX-like syndromes. These results imply that only a subset of IPEX-like patients could be eligible for cell therapy with in vitro expanded or gene-modified-Treg cells. Indeed, in depth immunological screening is necessary to assess whether the immunological disorder is or not Treg-dependent.

Lentiviral-mediated FOXP3 gene transfer into T cells. We successfully developed an efficient and reproducible protocol for the in vitro conversion of conventional CD4+ T cells into Treg-like cells by transduction with a bidirectional LV encoding for full length FOXP3 and ΔNGFR as marker gene for selection (LV-ΔNGFR-FOXP3). The protocol has been tested using CD4+ T cells from both healthy donors and 6 IPEX patients carrying different FOXP3 mutations. In all cases the resulting CD4FOXP3 Treg-like population maintains high and stable levels of FOXP3 protein expression, up-regulates Treg-associated markers (including CD25, CTLA4, Helios) and activation markers (CD69, CD71, HLA-DR). CD4FOXP3 T cells are hypoproliferative, acquire in vitro suppressive function, and suppress Th1 and Th2 cytokine production. High expression of FOXP3 and suppression of Th1 and Th2 cytokines is maintained even after in vitro culture in the presence of inflammatory cytokines, thus supporting the stability of the gene-modified cells (Passerini et al., Sci Transl Med 2013).
In order to make the use of CD4FOXP3-based T cell therapy safer and pave the way for the application of CD4FOXP3 Treg cell therapy beyond IPEX syndrome, we further challenged our LV platform by applying the bidirectional LV-FOXP3-ΔNGFR vector to achieve conversion of Ag-specific effector T cells into Treg cells. Specific protocols for the generation of nominal-Ags and Allo-Ags Ag-specific CD4FOXP3 T cells have been generated. Briefly, nominal Ag-specific CD4FOXP3 T cells were generated from CD8-depleted PBMCs activated in vitro in the presence of the nominal Ag, whereas allo-specific CD4FOXP3 T cells were produced from naïve CD4+ T cells stimulated with allogeneic LPS-activated mature DC. After 2 days, cells are exposed for 24h to concentrated LV: in these settings T cells activating in response to the Ag are preferentially transduced. Purification of the transduced T cells and further expansion in the presence of the cognate Ag allows the generation of Ag-specific CD4FOXP3 T cells which high FOXP3 and Treg-related markers, are specific for cognate Ag, are hypoproliferative and acquire the ability to suppress autologous T cell responses the their cognate allo-Ag. Their ability to maintain Ag-specific function in vivo is currently being tested in a humanized mouse model of Ag-specific response.

In summary, we demonstrated that bidirectional LV-mediated FOXP3 overexpression:
- is able to convert conventional CD4+ T cells into suppressive T cells, fully resembling Treg cells, and that this occurs regardless of the presence of FOXP3 mutations;
- is feasible in T cells from IPEX patients;
- can in principle be used for the generation of suppressor cells with known Ag-specificity.

Lentiviral-mediated FOXP3 gene transfer into haematopoietic or lymphoid precursors. In view of a HSC-mediated gene therapy for IPEX syndrome, we assessed, both in vitro and in vivo, whether LV-mediated FOXP3 over-expression in human HSC is feasible and safe. We first tested the effect of FOXP3 overexpression on HSC differentiation in vitro. Human UCB derived CD34+ cells were transduced using the bidirectional LV-FOXP3-ΔNGFR under a constitutive promoter. We have obtained on average 40% transduction and FOXP3 over-expression comparable to the levels of expression of human Treg cells. Transduced cells were: i) cultured in serum free and serum added liquid culture; ii) used for clonogenic assays in semi-solid medium; or iii) co-cultured on OP9-DL1 stromal cells (provided by Partner 2). Results showed that i) constitutive expression of FOXP3 in HSC improves maintenance of “stemness” markers and decreases proliferative potential; ii) the myeloid differentiating potential of HSC was not affected iii) upon T-cell polarizing conditions, FOXP3 overexpression resulted in a higher fraction of non-differentiating CD34+ cells and decreased CD4+/CD8+ ratio in differentiated cells. Moreover, FOXP3 overexpressing cells differentiated in higher proportion into CD4+CD25+FOXP3+ T cells. These data suggest that overexpression of FOXP3 in hematopoietic progenitors is not neutral. In order to confirm these findings, we assessed the effect of FOXP3 expression in HSC in vivo, by taking advantage of the humanized mouse model we have developed (see below). In vivo engraftment experiments with decreasing number of manipulated HSC confirmed that constitutive overexpression of FOXP3 in HSPCs leads to increased maintenance of “stemness” markers and decreases the differentiation of T cells. Indeed, FOXP3 transduced CD34+ cells contain a significantly higher frequency of SCID Repopulating Cells, when compared to control transduced cells, confirming that FOXP3 expression alters stemness-related networks in HSPCs. These results suggested that the construction of tissue-specific vectors to restrict the expression of FOXP3 in T cells may be needed for successful in vivo repopulation and differentiation of transduced HSC. To this aim, USR has generated a tissue-specific construct in which FOXP3 is under the control of human-CD4 promoter plus negative regulation of the murine CD4 silencer (which should avoid leaky expression in CD8+ cells) and the target sequence of miR223 (to avoid off-target expression in monocytes) (Brown Nat Biotech 2007, Marodon Blood 2003). This construct has been tested for its capacity to transduce human CD4+ T cells and to induce sustained transgene expression. Results obtained in healthy donor CD4+ cells showed low transgene expression by this construct and poor generation of Tregs in vitro. Previous data indicate that high levels of FOXP3 expression are required for the acquisition and maintenance of Treg features, thus suggesting that this construct may not be suitable for the generation of functional Treg-like T cells.

Pre-clinical animal models for the validation of in vitro generated Treg cells. In order to validate in vivo alternative therapeutic strategies for the cure of IPEX syndrome, Partner USR has developed humanized animal models to test the efficacy of in vitro generated Treg-like cells :

1. Xenogeneic graft-versus-host-disease (GvHD) model (in both NOD-SCID and NSG strains). We developed a humanized murine model of xenogeneic (xeno) GvHD using NSG mice and pre-conditioned NOD-SCID mice as recipients of human T cells. Lethal GvHD was induced by injection of allogeneic (allo) CD4+ Teff cells. CD4FOXP3 or control CD4NGFR T cells were co-injected or injected 6 days after Teff cells at a [1:1] ratio. Injection of control Teff cells in NSG mice induced xeno-GvHD development in 90% of animals. Co-injection of CD4NGFR T cells did not protect animals from disease, while CD4FOXP3 T cells injection, allowed the survival of 80 and 75% of mice, when injected at day 0 or later, at day 6, respectively. Co-transfer experiments in NOD-SCID hosts gave similar results, with 50% survival in mice receiving Teff cells together with CD4FOXP3 T cells. Late transfer of CD4FOXP3 T cells generated from IPEX peripheral CD4+ T cells efficiently protected animals from GvHD (75% survival), whereas transfer of CD4NFGR T cells did not. These results indicate that CD4FOXP3 T cells maintain are efficient suppressors also in vivo (Passerini et al., Sci Trensl Med 2013).
2. Transplantation of human UCB-derived CD34+ into neonate NSG mice. To allow full human hematopoietic differentiation in humanized mouse models we intrahepatically transplanted neonatal immunodeficient mice. We tested both the state of the art model NSG and the NSG-SGM3 model, which expresses human stem cell factor, granulocyte-macrophage colony forming factor and interleukin 3 and showed increased numbers of human T cells. Different conditioning regiments/treatments were tested and sublethal irradiation proved to be the most productive and reproducible. The model sustains human T cell differentiation with a distribution of the different stages of T differentiation similar to the human control in sub-lethally irradiated NSG mice. In the periphery, we detected both effector (naïve, effector, EM and CM) and Treg cells, in percentages similar to normal human donor controls. Therefore, we established that the intrahepatically transplanted sublethally irradiated NSG mice is a good model of in vivo human hematopoiesis.
3. Transplantation of human UCB-derived CD34+ carrying FOXP3 gene disruption or knock down (TALEN-mediated) into neonate NSG mice. In order to generate a humanized FOXP3-KO model, we envisaged the use of either shRNA or TALENs targeting the FOXP3 gene to knock-down or out FOXP3 in CD34+ UCB cells prior to transplantation in immunodeficient mice. Partner USR (in collaboration with M. Porteus, Stanford School of Medicine, CA) designed and selected a pair of TALENs efficiently disrupting the wt-FOXP3. Using an optimized nuclease-delivery protocol, we confirmed up to 20% FOXP3 knock-out in primary CD34+ cells, and 40% protein disruption in primary T cells. Upon transplantation of gene-modified HSC, T cell expansion was observed if efficient dysruption/knock-down was achieved (higher than 10% in BM cells). The phenotype includes accelerated thymic differentiation of conventional T cells and premature thymic regression, likely due to dampened thymic signal. This was associated with reduced naïve fractions in peripheral T cells and increased telomere shortening. Normal differentiation of Treg cells occurs (WT Treg cells outcompete FOXP3-KO/KD during differentiation). These data suggested that FOXP3 intrinsically controls early differentiation and peripheral homeostasis of conventional T cells and that its reduced expression triggers an immune-senescent phenotype. As alternative, in the course of the granted period we established a collaboration with the group of S. Snapper, Harvard Medical School, Boston (MA), who developed an improved humanized model for the reconstitution of the human hematopoietic system based on the use of NOD.PrkdcscidIl2rγ-/- (NSG) mice lacking murine major histocompatibility complex II and expressing human leukocyte antigen-DR1 (NSGAboDR1). We demonstrated that NSGAboDR1 can be used to generate a humanized FOXP3-KO model by neonatal intrahepatic transplantation of 30x10^4 CD34+ cells purified from the BM of an IPEX patient (reported in Goettel et al., Blood 2015). Correction of the altered immune phenotype in these mice might therefore be used as read out to assess the efficacy of CD4FOXP3 Treg cell therapy.
4. In vivo Ag-specific response. In order to assess whether CD4FOXP3 T cells interfere with Ag-specific responses, Ag-specific response was induced in NSG mice upon co-injection of PBMC and autologous Ag-pulsed (TT or Candida Albicans) monocyte-derived DCs. Two weeks after injection, mice were sacrificed and splenocytes were rechallenged in vitro with the same Ag, evaluating proliferation of the responder cells. Co-injection of autologous CD4FOXP3 T cells inhibited repopulation of human PBMC, suggesting their ability to inhibit the initiation of Ag-specific T cell responses while late injection of resting polyclonal autologous CD4FOXP3 T cells did not alter the magnitude of Ag-specific response, as measured upon in vitro re-challenge, suggesting that in the absence of inflammation/activation CD4FOXP3 T cells may not interfere with protective immune responses, thus hinting for safety of their in vivo use.

LV gene therapy for MUNC13-4 deficiency. Munc13-4 controls fusion of lytic granules with the plasma membrane in cytotoxic lymphocytes and its defect leads to impaired degranulation and cytotoxic activity in T and NK lymphocytes. Gene therapy could represent a therapeutic option for mutations in UNC13D gene encoding Munc13-4 protein resulting in type 3 of FHL (FHL3). INSERM partner constructed 4 integrative SIN-LVs in the pCCL backbone (Généthon) for Munc13-4 immunodeficiency. 2 vectors have been developed with the wide tropism VSV-G envelop for HSC transduction, and 2 vectors with T specific H/F envelop for T-cell transduction. Each vector carries the EF1α promoter driving codon-optimized Munc13-4 cDNA or a fusion between CFP protein and Munc13-4 as a reporter control. The complementation and restoration of defective phenotype of patient’s cells was first tested in vitro. INSERM partner has also performed similar experiments in vivo with the NSG mouse model, in collaboration with UCL, London. Injection of autologous gene-corrected EBV-cytotoxic T-cells (derived from a FHL3 patient) into the NSG mice bearing B-EBV lymphoma (generated from the same patient and injected into the mice) demonstrated the restoration of anti tumor activity of FHL3 T-cells. Gene-corrected FHL3 T-cells showed also higher persistence in PB in mice compared to non-corrected cells.
All together we demonstrated that the transduction of Munc13-4 deficient CTLs with our vectors leads to the expression of the transgene and the restoration of cytotoxic function as evidenced in-vitro as well as in vivo functional assays.
The in vivo proof of feasibility of HSC gene therapy was performed in a murine model of the FHL3 pathology, the Jinx mice witch has a mutation in Unc13d gene. The Jinx mice develop all signs of FHL when infected with LCMV. In this study we transduced Sca1+ HSC from the BM of Jinx mice with our vectors and engrafted them into the irradiated recipient Jinx. Four months post-transplantation the mice were infected with LCMV and FHL manifestation was monitored and compared between treated and non-treated groups. They observed reduction in HLH manifestation which correlated with a significant decrease of virus titer in the liver and serum level of IFN-γ and inflammatory cytokines. All these ameliorations might be explained by the restoration of cytotoxic function of CTLs as demonstrated in an in-vitro degranulation assay.
To improve transduction efficiency, a LV with Munc13-4 expressing backbone was pseudotyped with Measles virus glycoproteins H and F (H/F LV). Our comparative analysis between VSV-G and H/F vectors demonstrated that even at a low multiplicity of infection (MOI), H/F vector is more efficient to target activated FHL3 T-cells than VSV-G LV. This study makes H/F LV a good candidate in T-cell gene transfer, which needs to be further approved for clinical trials.

LV gene therapy for perforin deficiency. UCL partner generated and tested vectors for reconstitution of perforin deficiency using approaches based on HSC as well as T cell mediated gene therapy. They first investigated the correction of T cells derived from haematopoietic progenitors. In these experiments, they obtained lineage negative cells from perforin deficient mice and transduced them with vectors containing perforin. Cells were then cultured on stromal cells expressing signals for NK cell development. Differentiated cells transduced with the perforin vector were able to show restoration of cytotoxicity against targets cells in comparison to knockout or GFP corrected cells. Moreover they showed that the introduction of the correct copy of the perforin gene into HSCs can restore the T cell defects seen in perforin deficient mice. These mice are protected against development of HLH after viral challenge. Mice reconstituted with the perforin promoter showed higher expression in NK and CD8+ T cells compared to other cell types suggesting physiological transgene expression which did not interfere with other cell type differentiation and function,
These initial experiments provide a proof of principle for this gene therapy approach but it is known that HSCT into patients with active disease have a survival rate of only 50% and this may also affect the results of HSC gene therapy. Using the T cell mediated approach CD8 T cells were isolated from WT mice containing a CD45.1 background, expanded for 1 day and infected into sub-lethally irradiated perforin deficient mice (45.2 background). Twelve weeks after transplant CD8+ T cells were isolated, with these cells engraftment was analysed and cytotoxicity and IFN-γ production assays were performed. The results show that 1 month after transplant 50% of the CD8T cell population were derived from donor cells, this value decreased to 30% in the 2nd month and to 20% on the 3rd month. Nevertheless, 20% WT cells allowed full recovery of T cell cytotoxicity as well as reduction of IFN-γ production to levels very similar to WT mice. Similar results were obtained when naïve T cells were used in the transplant. The next step was to construct a retroviral vector to promote expression of perforin and GFP. This vector presented a transduction efficiency of T cells higher than 50% and this level of transduction was enough to allow recovery of in vitro cytotoxic activity in CD8+ T cells derived from perforin deficient mice. Overall, with these data and previously published studies reporting that it is possible to protect perforin deficient mice against HLH with only 20% correction, we believe that it is possible to control the manifestation of HLH through transplant of corrected effector cells. Thus, LV mediated perforin gene modification of autologous haematopoietic stem cells may be a therapeutic option for perforin deficient HLH.

Implementation of gene therapy trials for WAS and SCID-X1. Clinical trials have been implemented for SCID-X1 (UCL and INSERM see WP9), and WAS (UCL, INSERM, USR see WP10). Fourteen patients have been recruited for SCID-X1 and 18 patients for WAS (5 UCL, 5 INSERM, 8 USR) using similar gene transfer vectors. These protocols have been disseminated to sites outside Europe and 4 patients with SCID-X1 have been treated in US (Boston Childrens Hospital, Cincinnati Childrens Hospital, and UCLA), and 4 patients treated with WAS in Boston.

Adoption of other PID(s) for clinical study. Preclinical development has allowed adoption of 2 other PIDs for implementation of clinical protocols and others are close to enter clinical trials.
For ADA-deficient SCID we developed a self inactivating LV in which a codon optimised version of the human ADA gene is driven by an internal EFS (short form of the elongation factor-1alpha) promoter. A phase I/II trial was started to assess the safety and efficacy of EFS-ADA LV-mediated gene modification of autologous CD34+ cells from ADA-deficient individuals. Over 30 patients have been recruited between UCL and UCLA with excellent clinical outcome to date: good recovery of immunological function, including discontinuation of PEG-ADA and no adverse events. A manuscript describing interim safety and outcome is in preparation. Orphan drug designation has been gained in Europe, and a spin-out company from UCL has been created to support medicinal development (see below).
For X-CGD regulatory approval for a multicentre European trial sponsored by Genethon has recently been secured in London, Paris, Germany (GSH) and Switzerland, and a parallel study using the same vector and protocol established in 3 centres in US (Boston, Washington DC, Los Angeles). One patient in extremis has been treated using the specials exception clause of EU ATMP legislation, and one patient on trial. In two patients treated in London the vector enabled good levels of biochemical recovery in myeloid cells albeit transient, and early clinical improvement; unfortunately patients succumbed to complications relating to the severity of pre-existing disease. One patient has recently been treated in Boston with encouraging results at early time points (up to 6 months).
Clinical trials for ARTEMIS-SCID are in late-stage pre-clinical development including toxicology studies. Clinical trials for RAG2-deficient SCID are also in development.

Gene marking and safety evaluation. Analysis of clinical trials with SIN vectors (LV WAS, LV ADA-SCID and RV-SCID-X1) reveal polyclonal engraftment with no pathological clonal dominance. SIN vector configuration appears to abrogate integration site clustering observed in early LTR-based studies for SCID-X1 and WAS, suggesting that the mutagenic effect of vectors is substantially reduced in vivo (see WP9 and 10).

Long-term follow up of gene therapy trials with retroviral vectors and immunological monitoring.
Over 50 PID patients treated with gene therapy since 1999 are alive and followed long-term by 6 different clinical centers participating in the consortium and according to the study protocols and in line with requirements of national authorities and EMEA. Long term follow up on patients treated with g-RV demonstrate long term effectiveness of gene therapy for functional restoration of immunity using retroviral vector gene therapy (Gaspar et al. Science Transl Med 2011). In contrast to the leukemic event reported for CGD, WAS and SCID-X1, none of the ADA-SCID patients has developed complications due to insertional oncogenesis. we published an update on the ADA-SCID gene therapy trials using retroviral vector, investigating the long-term outcome of gene therapy in 18 patients with ADA-SCID. Overall survival was 100% over 2.3 to 13.4 years (median: 6.9 years). Gene-modified cells were stably present in multiple lineages throughout follow up. GT resulted in a sustained reduction in the severe infection rate from 1.17 events per person-year to 0.17 events per person-year. Immune reconstitution was demonstrated by normalization of T cell subsets (CD3+, CD4+, and CD8+), evidence of thymopoiesis, and sustained T cell proliferative capacity. B cell function was evidenced by immunoglobulin production, decreased intravenous immunoglobulin use, and antibody response after vaccination. All 18 patients reported infections as adverse events; infections of respiratory and gastrointestinal tracts were reported most frequently. No events indicative of leukemic transformation were reported. On April 1st the CHMP provided a positive opinion to Strimvelis, the drug name under which ADA-SCID RV gene therapy will be commercialised (see below).
Specialised immunological monitoring methodologies have been established in all clinical centres performing trials and are ongoing to closely monitor patients enrolled in clinical trials. In the context of the ADA-SCID trials using gammaretroviral vectors, through combination of immune phenotype and integration studies we provided high-resolution tracking of T cell fate and activity confirming in humans the safe and functional decade-long survival of engineered T-memory stem cells (Biasco et al., Science Transl. Medicine 2015), paving the way for their future application in clinical settings.

Previous clinical trials with a first generation gammaretrovirus vector expressing the IL-2 receptor gamma chain cDNA successfully restored immunity in children with SCID-X1, but resulted in vector-induced leukemia in 25% of subjects through enhancer mediated mutagenesis. We developed a SIN gammaretrovirus (SIN-γc) vector containing deletions in viral enhancer sequences to improve safety. Clinical trials have been implemented by INSERM (5 patients) and UCL (1 patient off protocol, 1 patient on protocol) (and also disseminated to Boston Childrens Hospital, 4 patients, UCLA 1 patient, and Cincinnati Childrens Hospital 2 patients) using identical vector and protocols. No adverse events have been noted in any patients to date. One patient has received allogeneic transplant, and one patient has been retreated as a result of poor engraftment of transduced cells.
The results on the first 9 patients were published in 2015 (Hacein-Bey-Abina et al., NEJM;). All patients received BM derived CD34+ cells transduced with the SIN-γc vector without preparative conditioning. After 12-38 months follow-up, 8 of 9 children are surviving. One patient died of overwhelming adenoviral infection prior to reconstitution with genetically modified T cells. Of the remaining 8 patients, 7 recovered PB T cells that were shown to be functional, thus leading to resolution of infections. Patients remained healthy thereafter. CD3+ T cell recovery was not significantly different from that observed in previous trials. These data show that a modified gammaretrovirus vector retains efficacy in treatment of SCID-X1, and suggests that outgrowth of T cell clones as a result of insertional mutagenesis is significantly abrogated. Analysis of integration sites in patients has progressed in collaboration with Dr Rick Bushman (Philadelphia). Early analysis has demonstrated polyclonal engraftment of T cells, and no pathological clonal dominance. As noted previously, comparison of integration sites using the SIN vector and LTR-based vectors (previous trials) revealed significantly less clustering of insertion sites within LMO2, MECOM and other lymphoid proto-oncogenes.
Currently, efforts are directed towards transferring the same expression cassette to an alternative LV-based platform, with planned trial initiation in early 2017.

Eighteen patients were treated in the 3 different centers, 16 were alive at last follow up and 1 patient treated at Necker has died due to pre-existing infectious disease complications. Four additional patients have been treated in Boston with the same LV. The investigational medicinal product (IMP) consists of autologous CD34+ HSC engineered with a LV driving the expression of WAS cDNA from an endogenous 1.6 kb human WAS promoter (LV-WAS). Cumulative results of both trials show multilineage engraftment of gene corrected cells, restored Wasp expression (preferentially in the lymphoid lineage), improved eczema, increased platelet counts, decreased infections and autoimmunity. These results lead to an improved clinical score and benefit for the patients. The data have been published in part by USR/FCSR in 2013 and a recent updated presented at the 2015 ASH meeting by F. Ferrua (oral communication). Another manuscript has been published by the UCL/NECKER patients describing the results (Hacein-Bey et al., JAMA, 2015).

Results of USR/FCSR clinical trial. The phase I/II clinical trial started in 2010 with gene therapy administered after a reduced intensity conditioning (RIC) based on anti-CD20 mAb,

targeted busulfan and fludarabine. (see Figure3) As of April 2016, eight patients have been treated at a median age of 2.2 years (range:1.1-12.4). They are all alive and well, with a median follow-up (FU) of 3.8 years (range:0.6-5.9). HSC source was BM, mobilized PB or both. Median IMP dose was 9.6x106 CD34+/kg, containing 88-100% transduced progenitors. Robust, persistent engraftment of LV-transduced cells was observed in BM and PB; WAS protein expression was restored in most platelets, monocytes and lymphocytes. After immune reconstitution, a marked reduction in severe infection rate was observed. Five patients stopped immunoglobulin supplementation, developing specific antibodies upon vaccination. Eczema resolved in all. Lymphocyte subset counts were normal in most patients and proliferative response to anti-CD3 mAb was in the normal range No clinical autoimmune manifestations were observed ≥1 year FU. After GT, all patients became platelet transfusion independent, with increasing platelet counts and markedly decreased severe-moderate bleeding frequency (See Figure 1). Neither IMP-related Serious Adverse Events nor evidence of abnormal clonal proliferations were observed. Quality of life improved in all patients after GT. From the 2nd year of FU, the number of hospitalizations for infections decreased and no hospitalizations due to bleeding were observed after treatment. In summary, TIGET-WAS data show that lentiviral GT for WAS is well tolerated and appears to lead to sustained clinical benefit. The high level of gene transfer obtained with LV-WAS results in robust engraftment of transduced HSC, even when combined with RIC. Two additional patients have recently been treated under hospital exemption. Prolonged FU will provide additional information on long-term safety and efficacy of this treatment.

Results of Genethon-UCL clinical trial. A recent published report from UCL/Genethon describes infusion of gene-corrected autologous HSCs in 7 patients with severe Wiskott-Aldrich syndrome lacking HLA antigen–matched related or unrelated HSC donors (age range, 0.8-15.5 years; mean, 7 years) following myeloablative conditioning. Patients were enrolled in France and England and treated between December 2010 and January 2014. Follow-up of patients in this intermediate analysis ranged from 9 to 42 months. Follow-up of patients in this intermediate analysis ranged from 9 to 42 months. Six of the 7 patients were alive at the time of last follow-up (mean and median follow-up, 28 months and 27 months, respectively) and showed sustained clinical benefit. One patient died 7 months after treatment of preexisting drug-resistant herpes virus infection. Eczema and susceptibility to infections resolved in all 6 patients. Autoimmunity improved in 5 of 5 patients. No severe bleeding episodes were recorded after treatment, and at last follow-up, all 6 surviving patients were free of blood product support and thrombopoietic agonists. Hospitalization days were reduced from a median of 25 days during the 2 years before treatment to a median of 0 days during the 2 years after treatment. All 6 surviving patients exhibited high-level, stable engraftment of functionally corrected lymphoid cells. The degree of myeloid cell engraftment and of platelet reconstitution correlated with the dose of gene-corrected cells administered. No evidence of vector-related toxicity was observed clinically or by molecular analysis. Additional patients were enrolled following this publication, for a total of 10 patients (5 Necker, 5 London).

Dissemination of non-classified information to a wide audience. CELL-PID discoveries were shared with scientists worldwide via presentations in international scientific conferences (e.g. ESGCT, ASGCT, ISSCR, ASH, ESID, EBMT) and publications in international peer-reviewed journals (e.g. Nature, Science, Scienec Transl Med., JAMA, Blood, JACI, Molecular Therapy, Human Gene Therapy, Gene Therapy). At the time of report finalization, 189 publications are posted in the Publication section of the Participant Portal and are advertised in the CELL-PID website open to the general public ( Most of the articles are “open access” to allow free and unrestricted access to these important publications. CELL-PID work and results are extensively presented at national and international conferences through oral presentations and poster presentations. A detailed list of conference presentations has been posted in “the dissemination activities” folder accessible from the Participant Portal and are attached to the final report.

Establishment of a mobility plan for scientists and key staff. Training of CELL-PID aimed to promote integration of different teams and education of new staff partly through the mobility of professors, students and technical staff. An annual mobility plan was implemented since the beginning of the project which involved overall 19 students/researchers from the CELL PID partners. These scientists spent from 5 days to 7 months in one of the CELL PID centers as part of training, learning techniques, or performing specific experiments.
Development of training programmes for scientific and regulatory personnel. Two CELL PID training course were held during the five year program which received overall a positive feedback from the participants. A training course on Gene Transfer in Human Hematopoietic Cells was held in Milano on 18-19 June 2012 in cooperation with USR, INSERM, MolMed and MILTENYI. The course was aimed at expanding the knowledge on hematopoietic cell gene transfer using integrating vectors, with particular regards to cell purification, transduction and characterization of the quality of the product. The course was attended by 17 participants from different Institutions and included both theoretical and practical sessions performed at USR labs and MolMed. On 5-6 June 2014, MHH in cooperation with the Fraunhofer Institute for Toxicology and Experimental Medicine (ITEM) hosted the first Workshop on Toxicology, bringing together basic researchers and scientists working on translation of medicinal products to the clinics. Members from four different Cell-PID partner organizations (EMC, UZH, USR and HU) attended the two day program for a total number of 12 trainees.
CELL PID participated to the organization and sponsored Clinical trial courses and Commercialization Workshops that were held at the 2015 ESGCT meeting, with participation of CELL PID partners. Topic included implementation of multi-center trial, new EU regulation, clinical Bio-Manufacturing Facility, setting up and running an academic cell factory, conducting a trial in an academic institution, and discussion on how balancing clinical needs with commercial realities.
Implementation of a communication infrastructure for the project. Annual Newsletters issued as an annual deliverable were posted into the CELL-PID website in the publication section). Scientific meeting were organized annually. They functioned as a major platform for presentations and exchanges between the PhD students and Post-docs working on the project, and as such be a major educational forum. International renowned speakers were invited to give presentations on their expertise and the major advances in their work. Annual meetings give the opportunity to organise annual Governing Boards meetings.
Exploitation and commercialisation. The most relevant exploitation and commercialization results are reported in the next section.

Contribution to EU standards. Several protocols and methods originating from the activities of different WP have been optimized, standardized and validated. These include methods to study HSC biology and lymphocytes, vector production, cell transduction, pre-clinical models, clonality and integration studies, safety and efficacy assays, and clinical follow up assays. The list of most relevant published protocols/methods/assays is described in deliverable 12.1 and the corresponding papers or documents have been uploaded in the portal.
Regulatory issues. For clinical regulatory aspects, a questionnaire was sent by USR clinical trial office to relevant CELL PID partners to obtain information on current clinical trial management in centres involved in the CELL-PID project and establish a network of the reference persons at the respective sites. Following the survey, it was confirmed that participating centers work in adherence to GCP and EU/international guidelines and have established procedures for running the clinical trials.
Partners consulted each other on specific issues related to new EU regulation for clinical trials. USR had responsibility to coordinate and supervise the activities aimed at collecting and verifying the necessary ethical and regulatory documentation for the CELL PID project. The USR office established strict contact with local clinical trial office and partners delegate to secure all the relevant documentation related to clinical trial authorization.
During the five year of the project members of CELL PID have participated to the following activities involving regulatory aspects in gene and cell therapy (see also above WP11):
1) clinical trial workshop and commercialization workshops held at ESGCT meetings, which brought together researchers and clinicians, from academia and industry, experts in regulatory affairs and members of national and EU regulatory authorities; 2) CELL PID training courses on “Gene Transfer in Human Hematopoietic Cells” and “New developments in the toxicology of gene-modified hematopoietic cells” organized at USR/FCSR and MHH respectively; 3) Stem Cell Clonality and Genome Stability Retreats hold in conjuction with ESGCT and ASGCT annual conferences.
Ethical issues. The coordinating team supervised throughout the project activities necessary to verify that CELL-PID clinical studies were adequately registered and approved, that the local authorities had approved animal experimentation and that the 3 R’s were respected.
The documentation related to preclinical and clinical studies was initially collected from the partners at the beginning of the project and uploaded in the portal. During the course of the project the Governing Board of CELL PID decided to delegate the task of verifying activities on ethical/regulatory aspects to the coordinator office. The CELL PID coordinator delegated to Dr. Zancan (USR), head of the USR clinical trial office, the responsibility to coordinate and supervise the activities aimed at collecting and verifying the necessary documentation for the CELL PID project. Partners and their delegates for ethical issues sent all necessary documents including approval of animal experimentation by the local authorities, ethical approval for the cell therapy in children, studies conducted on human biological samples. The documentation received was checked for compliance by the committee and queries for missing documents or unclear information were sent to partners as appropriate.
A database containing all the relevant ethical and regulatory documentation was then prepared. Documentation collected from partners (opinion, authorizations, informed consents) was divided according to: 1) Biological sample research; 2) Preclinical research; 3) Clinical trials; 4) Manufacturing of cells/vector; 5) Other activities.
A spreadsheet summarizing the documentation was prepared and all the documentation was uploaded on the partner-restricted section of the CELL PID web site.

Project Management. The coordinating team was assisted by a managerial partner (FINOVATIS), to ensure professional management of this complex and large-scale European research project. During the 5 years of the project, the management team at USR and FINOVATIS dedicated much of its time to setting up all actions, tools and systems necessary to ensure the good implementation of the project. This was done in order to ease partners’ work and avoid unnecessary administrative burden.
Annual meetings and partners interaction. Each year a CELL-PID meeting was held gathering delegates from the different partners. The scientific program dealt with most updated work from all scientific WP. There were usually one or two invited lectures from members of the advisory board and prominent scientists in the field of cell and gene therapy.
The governing board oversaw the progress of the project towards its objectives, deliverables and milestones, while ensuring the proper administrative, legal and financial status. Meetings of the governing board were held annually and several ad hoc meetings and TCs were organised.
IPR and knowledge. IPR/Knowledge management was based on the fundamental ethical rules and principles recognised at the European level, ensuring that CELL-PID Research is driven into the right directions. The access rights to pre-existing know-how and knowledge, were defined in the general conditions of the EC contract and detailed in the Consortium Agreement. Confidentiality of information and knowledge was enforced as defined in the Consortium Agreement. A list of novel IP and exploitable foreground is described in the report.
Gender actions. In order to promote the participation of women in this project and related research projects, a set of gender indicators was produced to measure progress towards gender equality in gene therapy research.

Potential Impact:
Primary immunodeficiencies (PIDs) constitute a large and heterogeneous group of rare heritable disorders resulting in an underdeveloped and/or functionally compromised immune system. World-wide, the incidence of PIDs varies greatly from 1 in 600 to 1 in 500,000 live newborns, depending both upon the specific disorder and the ethnicity of the population under study. Patients with PIDs display varied clinical phenotypes that can be life-threatening (for example various forms of SCID and severe myeloid disorders). For these conditions, allogeneic HSC transplant (HSCT) from a human leukocyte antigen (HLA)-matched donor confers lifelong therapeutic ‘cure’ with a success rate of more than 90% in the best centres1. Unfortunately, paucity of HLA-matched donors, particularly in non-caucasian populations, results in considerable increases in morbidity and mortality to varying extents depending on pre-existing clinical condition and genotype. For other less severe forms of PID, lifelong treatment with drugs may be only partially effective, can have major impact on quality of life, and may be limited by local economics or availability (for example immunoglobulin replacement in antibody-deficiency syndromes). These limitations are prominent in large populations (for example in China and South America), where healthcare systems are less well developed, and where autologous transplant procedures aimed at life-long cure would be particularly cost-effective.

The results of CELL PID project, most of which have been published in the open, peer reviewed scientific literature, justify the expectation that ex vivo lentiviral gene therapy of HSCs will become a highly efficient, cost-effective curative treatment modality for a variety of inherited disorders, with special reference to PIDs, other haematological disorders and lysosomal storage and metabolic disorders. In view of the very high costs of current, in most cases insufficiently effective treatment modalities for many of these inherited disorders the socio-economic impact cannot be overestimated. Results encompass basic biology studies, preclinical studies, technological platform and clinical outcome with high impact on the patient population.

This section describes the relevance and impact of the key results of the CELL PID project:

1. New information derived from studies on hematopoietic stem cells.
The ability to ex vivo expand of HSCs is of significant relevance for both autologous hematopoietic gene therapy in inherited disorders as well as for the frequently practiced allogeneic UCB transplantation in patients lacking and HLA-matched family donor. We obtained proof of principle for stem cell mobilization as an alternative for cytoreductive conditioning of the recipient patients. Moroever, we can favour HSC engraftment of ex vivo gene modified stem cells by improved homing and stimulating the proliferation rate of homing cells. Finally, a new immune deficient mouse strains can improve human cell engraftment capacity, enabling detailed evaluation of gene modified and expanded human HSCs.
Many of the findings of these studies are currently incorporated in the CliniMACS Prodigy® system aiming at a closed, GMP-compliant system for selection, transduction and expansion of the stem cells (See below, exploitation).

2. Thymus regeneration and transplantation
2.1 Studying thymic defects and exploring thymic regeneration. The analysis of thymic defect and its improvement after cytokine administration, HSC transplantation and gene therapy has allowed to dissect the cellular and molecular mechanisms involved in the pathogenesis of autoimmunity in Omenn syndrome. The observation that antiCD3 mab treatment induces thymic development and improvement of Omenn clinical signs in the mouse model, suggests that anti-CD3ε mAb might constitute a pre-transplantation treatment in immunodeficient patients in which lymphopenia is associated with poor thymus development and autoimmunity. In parallel, our data of KGF effect on the thymic epithelial component will allow to consider this compound in the pre-transplantation treatment, particularly in the gene therapy setting.
WP2 has established the conditions to grow thymic epithelial cells (TEC) in vitro whilst maintaining some of their physiological behavior including important transcripts and relevant phenotypic features. This advancement in ex vivo TEC growth paves the way for novel therapeutic opportunities designed for gene correction in TEC with a cell-autonomous monogenic defect that affects growth and function. This prospect will drive novel treatment strategies for a number of so far incurable immuno-pathologies. These findings although of immediate application, hold the promise to improve efforts to correct TEC-specific defects employing novel methodological approaches such as gene editing.

2.2 Thymus transplantation is the treatment of choice for DiGeorge syndrome. Studies in DiGeorge syndrome patients has helped establish thymus transplantation as the treatment of choice for DiGeorge syndrome. The marked increase in referrals in the last 12 months from the Paediatric Immunology community around Europe bears testimony to this. The fact there is a now a centre in Europe means patients will not need to travel anymore to USA for this treatment as happened in the past. The encouraging results of the studies on cryopreservation of tissue offer the promise of possible delivery of this treatment to the patient’s home country as well as offering the possibility of selecting donors or manipulating the thymus in order to reduce the rate of autoimmune complications. The work on identifying patients other than DiGeorge patients who might benefit from this treatment offers the possibility of a wider application of this technology.

2.3. In-vitro generated T cell precursors: a new way to fasten T cell reconstitution.
Based on results generated in the context of WP2 on T cell progenitors differentiating in vitro in T cells, INSERM is planning to implement a new phase I/II clinical trial in January 2017. The trial design is already defined and will include adult and pediatric patients affected by an inherited (including PID) or an acquired disease of the hematopoietic system requiring an HSC transplant in the absence of a genoidentical sibling donor or matched unrelated donor.
The clinical trial design will foresee enrolment of 22 patients that will receive both unmanipulated and DL-4 exposed CD34+ cells (ATMP). The primary objective is to test the toxicity of the ATMP. Secondary objectives are (i) the evaluation of the hematological and immunological and especially T cell reconstitution, and (ii) the number and severity of infections in the post-transplant period. This is an innovative approach for trying to accelerate de novo T-cell development bypassing the step of T-cell precursor delivery from the BM by adoptively transferring in-vitro generated T cell precursors. The latter should be able of immediately seeding the thymus and then rapidly generating a wave of donor derived, polyclonal, host-tolerant T-cells as we demonstrated in an in vivo transplantation model that mimic the human setting. Patients transplanted with a partially HLA compatible donor experience a long period of hospitalization, with an immunodeficiency can last up one year. During this period the prognosis of the patient is severe with a high risk of mortality and morbidity. After HSCT, de novo T cell generation from donor hematopoietic progenitors is impaired at different steps. First, the generation of T-cell progenitors in the BM and their delivery to the thymus has been shown to represent a limiting step in postHSCT recovery. Furthermore, conditioning regimen, graft versus host regimen infectious disease, and inflammatory status damage the thymic stroma and thereby alter intrathymic T cell differentiation.
For all these reasons, donor derived T cells generated in the recipient’s thymus first appear between 6 and 12 months after transplantation. With this procedure, we expect to significantly shorten the period of T cell lymphopenia without increasing the incidence of transplantation related toxicity. If we successfully shorten the length of the severe immuno-deficient period the benefit for the patients and public health is huge because: - The duration of hospitalization for each patient can be significantly reduced due to the reduction of the number and severity of opportunistic infections (ex: adenovirus, cytomegalovirus, entero viruses) - The use of partially incompatible HSC donors could thus be enlarged to all the patients who need this procedure even in transplantation units usually non familiar with this high risk procedure with a significant decrease of mortality and morbidity.

3. Alleviating the risk of insertional mutagenesis and standardising safety assays
Today, the risk of insertional mutagenesis prevents the broader use of gene therapy to treat more common diseases. CELL PID developed and evaluated new vector architectures which are currently used or will be used in upcoming clinical trials. The overall safety profile of the SIN-LVs most commonly used in the clinic has clearly been improved. The results from the treated patients will help to assess the real risks associated to the new vectors and, if they perform safely, eventually broaden the disease phenotypes which can be treated with gene therapy in the future. In principle, the newly developed molecular assay can also be used for cell types other than HSCs. It is conceivable that this assay could serve as a standard laboratory test during preclinical safety evaluation for gene therapy of different tissues, which is a currently unmet need. A better understanding of the mechanisms underlying the immortalization of transduced cells will enable us to not only prevent SAEs in the future, but to also transfer our knowledge of deregulated transcriptional pathways to similar, naturally occurring disease phenotypes. The identification of aberrant intracellular signalling in insertional mutants could potentially be used to identify inhibitors and drugs that are also effective in treating other proliferative disorders. These candidates could also be envisioned to serve as biomarkers in future clinical trials including, but not limited to, gene therapy. In conjunction with studies from our partners, our insertion site data can further be used to understand the hierarchical organization of the hematopoietic system and to help explore the nature of the real stem cells.

4. Development of HSC gene therapy for PID as a medical need
Development of HSCT gene therapy for PID represents a paradigm for translational research in many other rare diseases. The first successful allogeneic HSCTs were performed in patients with X-linked severe combined immunodeficiency (SCID-X1) and Wiskott-Aldrich Syndrome (WAS) in 1968. Thirty years later, major therapeutic benefits via gene transfer to autologous HSCs was first demonstrated in patients with SCID (X-linked and ADA-deficient), crucially benefitting from a major growth and survival advantage in vivo of corrected progeny. This strategy is now showing considerable promise in other disorders including WAS, Chronic Granulomatous Disorder (CGD), and other metabolic disorders, where in the absence of a substantial or consistent selective advantage, long-term correction necessitates the use of robust patient pre-conditioning. We have now treated over 50 patients with various forms of PID (SCID-X1, ADA-SCID, X-CGD, WAS), largely using vectors developed and evaluated pre-clinically in-house. Based on remarkably encouraging results from ongoing studies we are developing similar strategies for other tractable conditions.
Collectively, our data demonstrated that autologous gene therapy for PID can be conducted safely and with therapeutic efficacy at least comparable to mismatched HSCT.
Trials with larger numbers of patients will be required to confirm these observations and will require detailed retrospective analysis of data from both modalities of treatment, maybe even with randomized trials.

4.1 Gene therapy for SCID.
Today, 14 SCID-causing genes have been identified and their discovery has helped us to understand the pathophysiology of these life-threatening diseases, and has opened up new avenues for novel therapeutic approaches. Allogeneic HSCT from a healthy donor is associated with at least an 80% probability of survival with cure if the donor is an HLA-genoidentical sibling or a pheno-identical unrelated individual. In the absence of such a donor, use of a graft from a partially compatible family donor is associated with a 5-year survival rate of between 30% and 70%. This variability in survival depends on several parameters, such as the patient’s age, the presence of infectious complications at the time of transplantation, and the molecular cause of the disease.
We have been able thanks to the CELL PID network to better understand how gene therapy can potentially cure SCID. RAG2, ADA and Artemis projects are definitively the best examples and with our findings not only more patients can be enrolled into clinical trials but also new curative technologies could be developed in the near future.
Moreover, we were also able to demonstrate that gene therapy for some diseases such as RAG1 will require further efforts. Indeed, the finding that RAG1 endogenous expression could compete with RAG1 transgene expression explains many of the difficulties encountered within this project. We have now to consider this observation to design a LV that drives a high RAG1 transgene expression and eventually abrogate endogenous RAG1 expression using shRNA strategy, or seriously consider correction of the endogenous deficient RAG1 gene by a gene editing approach.
Finally, the socio-economic impact of gene therapy has to be more profoundly analyzed and taken into consideration also in comparison with allogeneic transplantation. All clues are indicating that not only survival, but also quality of life is largely improved by gene therapy as compared to classical allogeneic HSCT. If proven by extensive comparative studies, it will allow all patients who do not have an HLA matched-donor to benefit from a gene therapy.
Gene therapy for ADA-SCID will become a standard of care for some patients in the very near future, and it is likely that this will be true for other conditions over the next 5 years. This will inevitably require collaboration with pharma and biopharma in order to achieve global spread and market authorisation. There are examples of this emerging form CELL-PID including the licensing of a gammaretroviral gene therapy for ADA-SCID by GSK (collaboration with TIGET), and the creation of a spinout gene therapy company from UCLB London (Orchard Therapeutics). In economic terms, there has not yet been a formal evaluation but, from the experience so far, it would appear that an autologous gene therapy procedure is associated with much reduced hospitalisation costs. The added safety of new gene therapy protocols as well as the proven efficacy in several different conditions suggests that similar strategies can be successfully applied to many other PIDs as outlined in the proposal.

The Table below is from a recent review (Cicalese, Aiuti, 2015, Hum Gene Ther) and summarizes most of the ongoing clinical trials on primary immunodeficiencies in the world and related publications. Several CELL PID partners are leading these activities.

4.2. Gene therapy for CGD.
Until now the therapeutic options for CGD patients were mainly antimicrobial prophylaxis, early and aggressive treatment of infections and interferon gamma (IFN-g). Although HSC transplantation can cure CGD if an HLA-compatible donor is available, this approach is associated with significant morbidity and mortality. Gene therapy was seen as a unique treatment option for patients with CGD especially if no matched BM donor is available. Our work has turned this vision into reality. We succeeded in developing self-inactivating LVs targeting CGD which are highly specific, safe and efficient. The vectors developed during these studies successfully restored the underlying defect in CGD cells, namely the lack of superoxide production. Clinical trials with the G1XCGD vector have been approved in the UK, France, Switzerland and Germany and our knowledge and developments have been used to treat patients even outside of the EU. Gene therapy with the G1XCGD and MSP.gp91_2x126T vectors will certainly improve the life conditions of CGD patients. With this we have generated a novel therapy option for CGD patients which allow a single medical intervention rather than prolonged and often unsuccessful conventional symptomatic medical treatment as it was in the past. In addition, the development and application of vectors targeted to myeloid cells will make this the technology applicable for a wide range of patients independently of donor availability.
Several biotechnology companies (MOLMED, Miltenyi, Genosafe, GATC, Eufets, Plasmid Factory) were involved in our studies, thus creating a strong link between scientists involved in basic research and industry. This link will facilitate exploitation of our know-how to address the specific requirements and technical challenges raised during our studies. Furthermore our approach to direct therapeutic gene expression to myeloid cells avoids potential toxicity issues related to the ectopic expression of the transgene in cells were it is not normally expressed. This strategy will be also applicable for other diseases of myeloid cells and thus will paved the way for a broad application of gene modified stem cells for the treatment of disorders of the innate immune system. Our results will be important for future development and clinical application of a broad spectrum of stem cell based technologies targeting the myeloid compartment.

4.3 Gene therapy for IPEX and HLH.
Transplantation of HSCs is the only available treatment for FHL3 and IPEX, but can be ineffective and with complications. Gene therapy comes as the best curative solution for these patients. During the past five years, by working together to test the feasibilty safety and efficacy of a gene therapy approach, we have brought enourmous progress to the field. We have better clarified the disease pathology and gene function, established new in vivo models, demonstrated correction of defective function of very specialized T cell subsets. Our results have been disseminated throughout the scientific community and raised more interest and confidence in this curative approach. More fundings will provide further support to develop the gene therapy protocols that we have define to successful treatment.

4.4 Gene therapy for Wiskott-Aldrich syndrome.
WAS is a complex immune disorder of genetic origin characterised by recurrent infections, bleeding, atopy, auto-immunity and lymphoreticular malignancy. The high morbidity associated with HSC (HSC) transplantation from partially matched donors has prompted the search for alternative strategies based on gene-corrected autologous HSCs. HSC gene therapy could represent a valid alternative since it is applicable in principle to all patients, and has several advantages over allogeneic transplant, including the lack of graft versus host disease and reduced toxicity.
Based on the results of our clinical trials, LV gene therapy has become an effective and well tolerated therapeutic option in patients with severe WAS, and has potential to become a standard of care for patients without matched HSC donors. Moroever, these studies have provided key information on the feasibility of GMP production of patients’ cells, the minimal effective vector dose, the optimal HSC dose, and requirement for different preconditioning that will be essential to help the development of novel clinical trials in other PID as well as non PID diseases.

A key objective of CELL PID was to favour the clinical development of ATMP for primary immunodeficiencies, sharing knowledge and know how among partners and countries, and facilitating the interaction with industry and biotech to allow exploitation and commercialization of medicinal products.
As described below in details, CELL PID partners working within and across different WPs were successful in bringing forward innovative cell/gene therapies medicines achieving: i) development of automated systems for transduction of HSC, establishing new safety assays and methods to generate T cells ex vivo: ii) obtaining new orphan drug designation for 3 diseases (ADA-SCID, WAS, X-CGD); iii) licencing ex vivo HSC gene therapy product for PID through alliance with industry (San Raffaele/Telethon and GSK), or generating spin out with biotech (UCL and Orchard Therapeutics); iv) achieving EU market approval for the first stem cell gene therapy for a genetic disease (GSK and San Raffaele/Telethon).

5.1 Development of new methods and assay for gene and cell therapies.
5.1.1 Manufacturing of transduced cells in a close system. A close, automated stem cell transduction system was tested in the context of CELL PID to improve. Related patents for the “automated generation of genetically modified stem cells” have been obtained by Miltenyi (EP2937100A1, US20150307900 A1). According to Miltenyi, further research, development and verification/validation work is required for commercialisation
5.1.2 Genotoxicity assays. A patent has been submitted for a novel genotoxicity assay for integrating vector systems developed by MHH (WP3), for which they are working with an industrial partner towards a commercial product. This invention was the basis to apply for new funding from an organization especially focused on the reduction of animal experiments.
5.1.3 Ex vivo generation of T cells. In the context of WP2, a patent regarding the ex vivo generation of T-cell progenitors and its clinical applications in the field of cell and gene therapy has been submitted by INSERM.

5.2 . Orphan Drug Designation (ODD) of ATMPs
5.2.1 LV WAS gene therapy
a. On 6 June 2012, orphan designation (EU/3/12/998) was granted by the European Commission to Fondazione Telethon, Italy, for autologous CD34+ cells transfected with LV containing the Wiskott-Aldrich syndrome protein gene for the treatment of Wiskott-Aldrich syndrome. The sponsorship was transferred to GlaxoSmithKline Trading Services Limited, Ireland, in December 2014.
b. On 7 October 2013, orphan designation (EU/3/13/1196) was granted by the European Commission to Généthon, France, for autologous CD34+ cells transduced with a LV containing the human Wiskott-Aldrich syndrome gene for the treatment of Wiskott-Aldrich syndrome.
5.2.2 LV XCGD gene therapy. On 9 February 2012, orphan designation (EU/3/12/957) was granted by the European Commission to Généthon, France, for autologous haematopoietic cells genetically modified with a LV containing the human gp91(phox) gene for the treatment of X-linked chronic granulomatous disease
5.2.3 LV ADA-SCID gene therapy. On June 7, 2013, orphan designation (EU/3/13/1134) was granted by the European Commission to Prof. Bobby Gaspar, United Kingdom, (UCL) for “autologous CD34+ cells transduced with a LV containing the human ADA gene for the treatment of adenosine-deaminase-deficient severe combined immunodeficiency”.

5.3. Continuation of clinical development of ATMPs in the context of new EU projects.
During the course of the five year grant period, CELL PID partners worked on securing additional funds for clinical trials for projects stemming from CELL PID. Two networks were successfully funded by the EU: the NET4CGD (for XCGD gene therapy) and the SCIDNET (for SCID gene therapy).

5.4. ATMP commercialization through agreement and licensing with industry and biotech
5.4.1 LV WAS gene therapy licensed from San Raffaele/Telethon to GlaxoSmithKline (GSK). GSK opted to license WAS gene therapy, in November 2013 from Fondazione Telethon and Ospedale San Raffaele. The clinical study protocol, (TIGET-WAS; EudraCT number 2009-017349-77), is currently ongoing and is now sponsored by GSK. The study is conducted by OSR in the Paediatric Immuno-Haematology Operative Unit of San Raffaele Hospital.

5.4.2 UCL spin out of Orchard Therapeutics. Orchard Therapeutics is a biotechnology company incorporated in September 2015 founded by UCL scientists A. Thrasher, B. Gaspar, W. Quasim. Orchard employs a collaborative development model for its research programs, working closely with clinicians and researchers at leading academic centers and it has established formal partnership with UCL, Manchester, UCLA and Boston’s Childrens’ Hospital to develop ex-vivo lentiviral gene therapy in patients with ADA-SCID and other diseases.

5.5 EU Market approval of an ATMP for PID.
STRIMVELIS (ADA-SCID gene therapy) EU approval. On April 1st the CHMP provided a positive opinion to Strimvelis, the drug name under which ADA-SCID RV gene therapy will be commercialised. EU marketing authorisation was granted on May 27, 2016. The decision was based on data collected from 18 children treated with Strimvelis. A 100% survival rate at three years post-treatment with Strimvelis was observed for all children in the pivotal study (n=12) and every child receiving the treatment who contributed to the marketing authorisation data package is alive today (n=18), with a median follow-up duration of approximately seven years. RV-gene therapy for ADA-SCID was originally developed by Ospedale San Raffaele and Fondazione Telethon (at TIGET) and was taken forward by GlaxoSmithKline (GSK) to the market through a strategic collaboration formed in 2010 between GSK, OSR and Telethon. Strimvelis is indicated for the treatment of patients with severe combined immunodeficiency due to adenosine deaminase deficiency (ADA-SCID), for whom no suitable human leukocyte antigen (HLA)-matched related stem cell donor is available.

These results highlight the importance of a EU network to facilitate interaction, sharing of information and know how among different partners. It also shows how industry and biotech are beginning to enter into the arena of rare diseases and ATMPs and cooperate with academia and non profit institutions towards commercialization of new medicines.

List of Websites: