ADA2 deficiency: from the molecular diagnosis to the pathogenesis and gene therapy

Background. Primary immunodeficiencies (PIDs) represent a heterogeneous group of monogenic disorders which are part of inborn errors of immunity characterized by immune dysregulation, altered immune responses of innate and/or adaptive immunity, and infections. More than 406 distinct disorders have been identified, resulting from variants in over 430 genes. Their number is rapidly increasing thanks to the application of next-generation sequencing (NGS), allowing the diagnosis in 15–70% of PID clinical cases. Among the recently identified conditions, adenosine deaminase 2 deficiency (DADA2) is a rare inherited disorder characterized by a wide variety of clinical manifestations and different ages of onset. The principal clinical features consist of vasculitis, early-onset ischemic strokes, intracranial hemorrhages, hematologic abnormalities (pure red aplasia), and immunodeficiency (neutropenia and decreased mature B cells). The ADA2 gene encodes an adenosine deaminase protein mainly secreted by macrophages. DADA2 pathogenesis is still not well understood. Biological anti-TNF therapy controls strokes and inflammatory episodes, while allogeneic hematopoietic stem progenitor cell (HSPC) transplantation is the treatment of choice for patients with hematologic manifestations. However, the morbidity and mortality associated with this procedure make this treatment only recommended for the most severe patients, and compatible donors are not always available. Hence, gene therapy based on autologous transplantation of genetically corrected HSPCs may represent a valid therapeutic option for treating DADA2 patients who are not eligible for allogeneic transplantation. Methods. For the project on diagnosis of PIDs, we designed an Haloplex gene enrichment target system including more than 630 genes among PID and candidate genes to investigate causative genes in fifteen patients clustered in T-cell defects, humoral defects, and complex PIDs main categories. Bioinformatic analysis was performed to obtain a molecular diagnosis for our PID patient cohort. For the DADA2 project, we characterized the DADA2 patient’s macrophage polarization defects using quantitative RT-PCR and ELISA. We generated an ADA2-deficient myelomonocytic U937 cell line to study macrophage defects associated with loss of ADA2. In particular, we evaluated the type-I IFN responses to exogenous double-stranded DNA and the involvement of the NLRP3 inflammasome in the increased pyroptosis susceptibility. In addition, we generated a third-generation lentiviral vector (LV) encoding human ADA2 under a ubiquitous promoter. The LV-ADA2 was used to transduce mouse and human HSPCs and assess transduction efficiency, clonogenic potential, and growth rate of ADA2-transduced cells. The ability of ADA2 transduced HSPCs to engraft in vivo was evaluated in humanized NSGW41 and C57BL/6 mice. Transduction efficiency and ADA2 reconstitution were also examined in bone marrow-derived CD34+ HSPCs isolated from two adult patients with DADA2. We also evaluated the correction of cytokine hyperproduction in patients’ macrophages and the ADA2-deficient myelomonocytic U937 cell line. Results. Using targeted NGS approaches, we achieved a molecular diagnosis for 16% of total screened patients, which comprised 23.5% of T-cell defects and 26.6% of the other PID patients. Four patients presenting with T-cell defects harbored mutations in ADA, IL7R, RAG1, and ARPC1B genes. Four patients manifesting complex clinical manifetations were diagnosed with JAGN1, DADA2, NRAS mutations, and one patient displaying homozygous variants in both MYD88 and CARD9 genes. Accurate clinical and immunological characterization was important to facilitate molecular diagnosis, indeed confirmed by NGS. Being DADA2 a newly identified disease with unclear pathogenesis, we investigated DADA2-associated defects and developed a pre-clinical GT approach mediated by a lentiviral vector (LV) encoding human ADA2 (LV-ADA2). Patients’ and ADA-deficient macrophages presented an unbalanced M1/M2 polarization ratio, which could be rescued by restoring ADA2 using the LV-ADA2, leading to a complete correction of the exaggerated cytokine response. ADA2-deficient macrophages produced increased IFNb levels in response to foreign dsDNA, suggesting an IFN signature as part of the DADA2 manifestations. ADA2-deficient macrophages also presented an increased cell death by pyroptosis upon NLRP3 inflammasome activation, indicating that inflammasome could exacerbate patients’ inflammatory condition. Transduction efficiency evaluated on individual vector-positive granulocyte-macrophage colony-forming units (CFU-GM) and erythroid burst-forming units (BFU-E) was > 75% in GFP and ADA2 transduced HSPCs. The median vector copy number measured in CFU-GM and BFU-E colonies ranged 1.82-2.69 and 1.58-3.7, respectively, at all vector concentrations. HSPCs CD34+ cell transduction with the LV-ADA2 led to a supra-normal increase of ADA2 expression and enzymatic activity in healthy donors and patients. LV transduction of patients’ CD34+ cells did not affect their clonogenic potential and differentiation capacity. ADA2 expressing human and mouse HSPCs engrafted and normally differentiated in mice. Conclusions. Using an NGS approach, we identified a CID patient manifesting WASlike signs associated with ARPC1B deficiency and other PID patient displaying typical manifestations associated with DADA2. These results confirm the power of NGS technologies detecting causative gene mutations in typical and complex PID phenotypes. Early and specific understanding of the genetic and immune defects of new PIDs will significantly contribute to finding new diagnostics markers and developing tailored therapeutic strategies as an alternative to allogenic transplantation. It was found that DADA2 patients are characterized by an unbalance M1/M2 ratio characterized by increased M1 and defective M2 polarization. ADA2-deficient macrophages also presented IFN signature and an increased cell death susceptibility in response to NLRP3 inflammasome activators. LV-mediated ADA2 reconstitution in patients' M1 and ADA2-deficient U937 macrophages restored a physiological secretion of TNF and IL-6. HSPC transduction with the LV-ADA2 led to a dose-dependent increase of intracellular expression and release of ADA2. Transgenic ADA2 exhibited a proper enzymatic activity. The ADA2 overexpression did not affect the HSPCs clonogenic capability of healthy donors and DADA2 patients' CD34+ cells. In both humanized and wild-type mice models, the engraftment of ADA2-transduced HSPCs was efficiently achieved, human and mouse HSPCs were capable to generate an hematopoietic system without remarks of distress or transduction-related toxicity. Our data support LV-mediated gene transfer in HSPCs as a new advanced therapy for patients with ADA2 deficiency and warrant further studies in support of its clinical development.