WAS Promoter-Driven Lentiviral Vectors Mimic Closely the Lopsided WASP Expression during Megakaryocytic Differentiation

Original Article WAS Promoter-Driven Lentiviral Vectors Mimic Closely the Lopsided WASP Expression during Megakaryocytic Differentiation Pilar Muñoz, 1,2,5 María Tristán-Manzano, 1,5 Almudena Sánchez-Gilabert, 1,5,6 Giorgia Santilli, 2 Anne Galy, 3,4 Adrian J. Thrasher, 2 and Francisco Martin 1 1 Genomic Medicine Department, GENYO, Centre for Genomics and Oncological Research, Pfizer-University of Granada-Andalusian Regional Government, Parque Tecnológico Ciencias de la Salud (PTS), Avenida de la Ilustracion 114, 18016 Granada, Spain; 2 University College London (UCL) Great Ormond Street Institute of Child Health (ICH), 30 Guilford Street, WC1N 1EH London, UK; 3 Genethon, 91000 Evry, France; 4 Université Paris-Saclay, Univ Evry, Inserm, Genethon, Integrare research unit UMR_S951, 91000 Evry, France Transplant of gene-modified autologous hematopoietic progenitors cells has emerged as a new therapeutic approach for Wiskott-Aldrich syndrome (WAS), a primary immunodeficiency with microthrombocytopenia and abnormal lymphoid and myeloid functions. Despite the clinical benefits obtained in ongoing clinical trials, platelet restoration is suboptimal. The incomplete restoration of platelets in these patients can be explained either by a low number of corrected cells or by insufficient or inadequate WASP expression during megakaryocyte differentiation and/or in platelets. We therefore used in vitro models to study the endogenous WASP expression pattern during megakaryocytic differentiation and compared it with the expression profiles achieved by different therapeutic lentiviral vectors (LVs) driving WAS cDNA through different regions of the WAS promoter. Our data showed that all WAS promoter-driven LVs mimic very closely the endogenous WAS expression kinetic during megakaryocytic differentiation. However, LVs harboring the full-length (1.6-kb) WAS-proximal promoter (WW1.6) or a combination of the WAS alternative and proximal promoters (named AW) had the best behavior. Finally, all WAS-driven LVs restored the WAS knockout (WASKO) mice phenotype and functional defects of hematopoietic stem and progenitor cells (HSPCs) from a WAS patient with similar efficiency. In summary, our data back up the use of WW1.6 and AW LVs as physiological gene transfer tools for WAS therapy. INTRODUCTION Wiskott-Aldrich syndrome (WAS) is an X-linked rare primary immunodeficiency (incidence of 1 in 10 5 to 1 in 10 6 cases per live birth) (OMIM: 301000) and makes up approximately 3% of all primary immunodeficiency disorders. It is diagnosed early in life, and many patients with severe WAS do not survive past 10 years of age without definitive treatment. 1 The classical WAS phenotype is characterized by eczema, immunodeficiency, microthrombocytopenia, autoimmunity, and malignancies. 2 WAS is caused by mutations in the WAS gene 3 (gene map locus Xp11.23-p11.22) that are expressed exclusively in hematopoietic cells and play important roles in signaling and actin cytoskeleton reorganization (reviewed in Blundell et al. 4 ). Therefore, most hematopoietic cells are affected to different degrees, causing the phenotypic abnormalities observed in patients with WAS. Of all functional defects, increased bleeding due to low platelet (PLT) counts is the major challenge, with up to 30% of WAS patients suffering lifethreatening bleeding episodes. 5 The only curative treatments for WAS patients are allogeneic hematopoietic stem and progenitor cell (HSPC) transplantation 6 and autologous gene-modified HSPCs. 7 Although HSPC transplantation is the standard treatment procedure and is usually curative, the use of human leukocyte antigen (HLA)-matched HSPCs is associated with acute morbidity and a high incidence of long-term complications, although overall survival rates are continually improving. 8–10 On the contrary, lentiviral-based gene therapy (GT) clinical trials observed limited toxicity and similar survival rates, 7,11–14 making this treatment a rational alternative therapeutic option. In these trials, autologous CD34 + cells were genetically modified using a self-inactivating (SIN) lentiviral vector (LV) expressing WASP under a 1.6-kb fragment of the proximal promoter of the WAS gene. 15 A general consensus that arose from these clinical trials is that, although immune deficiency was corrected, complete remission of microthrombocytopenia was more difficult to achieve. 16 Of note, HSPC transplantation is more effective than GT in this aspect and is therefore a drawback to overcome in new WAS GT products. 16 Received 6 May 2020; accepted 11 September 2020; https://doi.org/10.1016/j.omtm.2020.09.006. 5 These authors contributed equally to this work. 6 Present address: Karuna Good Cells Technologies SL, C/Cercas Bajas, 13-BJ, 01001 Vitoria-Gasteiz, Álava, Spain. Correspondence: Francisco Martin, Genomic Medicine Department, GENYO, Centre for Genomics and Oncological Research, Pfizer-University of GranadaAndalusian Regional Government, Parque Tecnológico Ciencias de la Salud (PTS), Avenida de la Ilustración 114, 18016 Granada, Spain. E-mail: [email protected] 220 Molecular Therapy: Methods & Clinical Development Vol. 19 December 2020 ª2020 The Author(s). This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/). The reasons behind the low PLT recovery achieved with GT in most WAS patients are unclear. It could be due to the absence of the selective advantage of WASP-expressing PLTs, 17 to the suboptimal WASP expression in megakaryocytes (MKs), MK progenitors, and/or PLTs, 14 or to a combination of both factors. In this direction, our hypothesis is that a physiological expression of WASP would improve GT outcomes by preventing potential side effects of underexpression or overexpression of WASP along MK differentiation. The exact functions of WASP in PLTs remains largely unknown, but there is strong evidence suggesting that it plays a critical role regulating MK differentiation and PLT formation by inhibiting these processes in the absence of the appropriated signals. 18–20 This could sound contradictory with the thrombocytopenia found in WAS patients, but it is actually a potential explanation for it. Indeed, PLTs produced in the absence of WASP have not been developed properly and have a reduced size, abnormal ultrastructure, and surface markers that lead to their elimination in the spleen and other tissues. Therefore, in order to generate normal PLT levels in WAS patients, we should not only engraft with the appropriate levels of WASP-expressing HSPCSs, but also mimic very closely the physiological expression of WASP during MK development. Different groups, including ours, have developed physiologically regulated LVs for the treatment of WAS using different fragments of the WAS-proximal promoter to drive the expression of WAS cDNA. 15,21 These studies have shown hematopoietic-specific expression of the different WAS promoter-driven LVs that efficiently restored WASP defects in animal models. 22,23 Later studies also showed improved safety of these LVs by avoiding WASP expression in non-hematopoietic cells 24 and reducing genotoxicity. 25 These works led to the approval of the clinical trials mentioned above using the 1.6-kb WAS-proximal promoter-driven LVs. 15 However, as referred to previously, despite the good clinical results, these trials showed suboptimal WASP expression in PLTs and suboptimal recovery of microthrombocytopenia in most patients. In an attempt to improve the behavior of WAS-proximal promoter-driven LVs, our group generated AWE LVs 26 that harbored regulatory fragments from the two WAS promoters described in the literature, i.e., the proximal promoter 27 and the alternative promoter. 28 The proximal promoter starts immediately upstream of the transcription start site (GenBank: AC115618.3 sequence [Seq] 120189–121855), and the alternative promoter is located 6 kb upstream (GenBank: AC115618.3 Seq 114403–115000). The AWE LVs showed improved enhanced GFP (EGFP) expression in myeloid, megakaryocytic, and B cell lineages compared to WAS-proximal promoter-driven LVs. 26 However, since the WAS cDNA contains regulatory sequences that affect its expression pattern, 29 it is fundamental to study the behavior of the backbone of AW LVs expressing WAS cDNA in order to determine whether they achieve truly physiological expression. In this work we use in vitro models to study whether WAS promoter-driven LVs harboring sequences from the alternative promoter could improve their therapeutic potential by mimicking more closely the WASP expression pattern during megakaryocytic differentiation. Although WASP expression levels are well documented in all mature blood cells, 30,31 the precise WASP expression profile in HSPCs along MK differentiation and PLT generation is mostly unknown. We therefore analyzed first the WASP expression kinetic during MK differentiation and then compared it with the expression pattern of the different LVs driving the transgene through different regions of the WAS promoter, including the LVs used in ongoing clinical trials for WAS. Our data showed that all WAS promoter-driven LVs mimic very closely the WASP endogenous expression kinetic during MK differentiation. We also studied potential therapeutic improvements of the AW LVs compared with the WW 1.6-kb (WW1.6) LVs, but could not find any significant differences. RESULTS WASP Endogenous Expression Decreases during In Vitro Megakaryopoiesis and Thrombopoiesis Since we wanted to mimic WASP endogenous expression during MK differentiation, we first established primary (Figure 1) and immortalized (Figure 2) cellular models to study WASP expression patterns through the MK differentiation process. MK progenitors (CD34 + CD41 + ), MKs (CD34  CD41 + CD42 + ), and PLTs (CD41 + CD42 + FSC low SSC low ) were routinely obtained from HSPCs (CD34 + CD41  CD42  ) using the protocol depicted in Figure S1A, where human CD34 + cells were incubated with stem cell factor (SCF), thrombopoietin (TPO), and ROCK inhibitor as described in Materials and Methods during 16 days. We could observe large cells that correlate with immature MKs (Figure 1A, center) that finally extend proplatelets (Figure 1A, bottom) and are also associated with the phenotypical acquisition of CD41a and CD42b megakaryocytic markers (Figure 1B). In addition, PLTs derived from HSPCs were functional in response to thrombin (Figures 1C and 1D) and expressed similar WASP protein levels as did those PLTs isolated from peripheral blood of healthy donors (HDs) (Figure 1E), validating this cellular model to study MK differentiation. We then analyzed WASP expression levels (Figures 1F–1H) in MK progenitors, MKs, and PLTs at days 4, 8, 11, and 16 of MK differentiation following gate strategy analysis represented in Figure S1B. Briefly, we first selected three different gates according to forward scatter (FSC) and side scatter (SSC) (HPCs, MKs, and PLTs), and then we further gated (1) undifferentiated CD34 + cells; (2) MK early progenitors (CD34 + CD41 + cells); (3) MK progenitors (SSC high CD34 + CD41 + ); (4) MKs (CD41 + CD42 + SSC high ); and (5) PLTs (CD41 + CD42 + SSC low ) as populations of interest. WASP expression was calculated as the intensity ratio of WASP + cells and the background intensity of the isotype control of the selected population (Figure S1B; see Materials and Methods). Our data showed a clear WASP downregulation as the HSPCs differentiate to MK progenitors (Figures 1F–1H; CD34 + versus CD34 + CD41 + , respectively), a slight WASP increment when MK progenitors (CD34 + CD41 + SSC high ) differentiate into early MKs (CD34 + CD41 + CD42 + , day 4), and a gradual downregulation as the MKs mature into PLTs (days 8, 11, and 16 and PLTs). The percentage of WASP + cells showed the same described tendency but in a more dramatic manner. We then validated these results in K562 and MEG-01 cell lines previously characterized to give rise to MK-like cells (Figures 2A and 2B) as determined by the acquisition of CD41a and CD42b megakaryocytic www.moleculartherapy.org Molecular Therapy: Methods & Clinical Development Vol. 19 December 2020 221 (legend on next page) Molecular Therapy: Methods & Clinical Development 222 Molecular Therapy: Methods & Clinical Development Vol. 19 December 2020 markers (Figures 2C and 2E). K562 cells differentiated with phorbol myristate acetate (PMA) and MEG-01 cells with valproic acid (VPA) constitute classical models to study megakaryocytic differentiation, which partially mimic some megakaryocytic characteristics in terms of phenotype, maturation, and function. 32 Interestingly, WASP expression levels and percentage of positive cells were increased upon MK differentiation in K562 cells (Figure 2D), while MEG-01 cells showed a clear downregulation along differentiation time (Figure 2F). These data indicate that K562 cells behave similar to MK progenitors that differentiate into early MKs in the presence of PMA, while MEG-01 cells resemble early MKs that give rise to mature MKs and PLTs upon the addition of VPA. 32 These two cell lines therefore provide additional cellular models to study LV behavior during MK differentiation. WAS Promoter-Driven LVs Mimic WASP Expression Kinetics during MK Differentiation and Restored PLT Actin Nodule Formation Once we established the three different human models of MK differentiation and their WASP expression kinetics, we proceeded to analyze the behavior of different second-generation LV backbones expressing the EGFP reporter gene through different fragments of the WAS promoter (Figures 3AandS2): the WE LV 21 harbors the 500-bp core of the WAS proximal promoter, and the AWE 26 and the cAWE (this work) contained additional sequences from the WAS alternative promoter. To analyze the physiological expression of these LVs during MK differentiation (Figure 3B), K562 cells (Figures 3CandS3A), MEG-01 cells (Figures 3DandS3B), and HSPCs (Figures 3EandS4) were transduced with the different LVs at a multiplicity of infection (MOI) of 1 for K562 and MEG.01 cells and at an MOI of 50 for HSPCs, obtaining a similar efficacy of transduction (Figure S5). Their EGFP expression kinetics (how the expression changes related to the expression of non-differentiated cells, referred as fold expression and detailed in Figures S3 and S4) were compared to that of endogenous WASP (filled bars in Figures 3C–3E) before and after MK differentiation. We observed similar behaviors of the three LVs, and all followed WASP expression kinetics upon MK differentiation in K562 cells (Figure 3C; upregulation), MEG-01 cells (Figure 3D; downregulation), and HSPCs (Figure 3E; downregulation). Interestingly, in the HSPC model, the kinetics of transgene expression at different times of MK differentiation were also very similar (Figure 3E). We next analyzed whether the physiological expression was maintained in the different WAS promoter-driven LVs expressing the WAS cDNA, i.e., WW, AWW, and cAWW LVs (Figure 4A). In this study, we used homozygous WASKO-K562 cells previously generated by our laboratory 20 (Figures 4B–4D) and WAS-null HSPCs from a WAS patient (Figures 4F–4I). WASKO-K562 cells were transduced with the different therapeutic LVs at an MOI of 1 to reach similar transduction efficacies (as shown in Figure S5). WASP expression levels were analyzed in different populations (see Figure S6 for details) after MK differentiation. Similarly to EGFP LVs, all WASP LVs increased WASP expression upon MK differentiation, measured as the percentage of WASP + cells (Figure 4C) and as Median of Fluorescence Intensity (MeFI) (Figure 4D). However, the WW LVs showed higher WASP expression in resting K562 cells (Figure 4D, WW) compared to endogenous WASP (Figure 4D, wild-type [WT]), AWW, and cAWW LVs, indicating a more physiological behavior of the AWW and cAWW LVs. Interestingly, AWWand cAWW-transduced WAS patient HSPCs also recovered a WASP expression kinetic more similar to HD HSPCs compared to those transduced with the WW LVs (Figures 4E and S7). Indeed, although the expression levels of WW LVs in undifferentiated CD34 + were closer to those of HDs compared to AWW and cAWW (Figures 4E and 4F), the expression kinetic at early days of differentiation (day 4) were different (Figure 4E, red line). Nevertheless, WASP expression levels in PLTs derived from WAS patient HSPCs transduced with the different LVs (MOI of 50) were similarly rescued (Figure 4F). In agreement with WASP expression levels, the restoration of actin nodule 33 formation in contact to immobilized fibrinogen, which is severely compromised in patient PLTs, was also similar for all LVs (Figures 4G and 4H). Phenotypic Correction of WASKO Mice after Transplantation with WWand AWW-Transduced WASKO Murine HSPCs (mHSPCs) We next analyzed the therapeutic efficacy of AWW and WW LVs in a mouse model of WAS, 34 because cAWW LVs did not show significant improvement in vitro compared with the AWW LVs. Lineage (Lin)  cells were isolated from bone marrow of WASKO mice (WASKO mHSPCs) (see Materials and Methods) and transduced with AWW and WW LVs at an MOI of 100. Both LVs achieved similar Figure 1. WASP Endogenous Expression Decreases during In Vitro Megakaryopoiesis and Thrombopoiesis (A) Representative bright-field images of hHSPCs before (top panel, 0–4 days) and after 4–8 days (middle panel, MKs are indicated with white arrows) or 11–16 days (bottom panel, MKs releasing proplatelets, white arrows) of MK differentiation with TPO and SCF without iROCK as indicated in Materials and Methods. (B) Representative plots showing changes in morphology (FSC high SSC high ) and appearance of megakaryocytic markers (CD41a and CD42b) of HSPCs cultured with the MK differentiation media. (C) Mature MKs (FSC high SSC high CD42 + cells) respond to thrombin by increasing intracellular calcium uptake (n= 3). (D) Expression of PAC-1 activation marker in response to thrombin in obtained MKs (FSC high SSC high CD42 + cells) and PLTs (FSC low SSC low CD42 + cells) (n = 3). (E) Representative histograms showing WASP expression levels of PLTs (FSC low SSC low CD41 + CD42 + ) obtained from peripheral blood (PB, top left) and HSPCs (bottom left). WASP expression (related to isotype control [IsC]) (right) is represented as mean ±SEM (non-parametric Mann-Whitney test, two-tailed). (F) Representative dot plots of endogenous WASP expression kinetics upon MK differentiation in vitro. Grey populations in each plot display the IsC staining and colored populations display the WASP staining (n = 7). (G) Analysis of WASP expression, depicted as MeFI of WASP + population/MeFI of IsC, in the different populations during MK differentiation. CD34 + indicates HSPCs at day 0 (CD34 + CD41  CD42  ), MK progenitors (CD34 + CD41 +- SSC low SSC high ) at days 4–8, mature MKs (CD34 + CD41 + CD42 + ) during 4–16 days of differentiation, and PLTs (CD41 + CD42 + FSC low SSC low ) obtained at day 16. (H)Percentage of WASP + cells in the above-described populations. Represented data are mean ±SEM. *p < 0.05, **p < 0.01, ***p < 0.001 (non-parametric Mann-Whitney test, two-tailed, compared to CD34 + expression levels) (n = 7). www.moleculartherapy.org Molecular Therapy: Methods & Clinical Development Vol. 19 December 2020 223 transduction efficiencies (Lin  AWW = 9.8 vector copy numbers/cell [vcn/c]; Lin  WW = 8.2 vcn/c) and expressed similar WASP levels in WASKO mHSPCs (Figure S8B) and in their myeloid progeny (Figures S8E and S8F). Lethallyirradiated WASKO mice (9.5 Gy) werethen injected with 3 10 5 AWW, WW, and mock WASKO mHSPCs as well as with WT mHSPCs (C57BL/6 mice). After 2 and 6 months of transplant, the SSC and FSC of cells from peripheral blood of the transplanted mice were analyzed (Figure 5A, left). The non-transduced transplanted mice showed a higher number of granulocytes and lower number of lymphocytes; meanwhile, the WT, AWW, and WW0.5 corrected mice showed a normal distribution of the population in peripheral blood. 6 months later, treated mice were sacrificed in order to analyze the ability of WW and AWW LVs to restore phenotypic and functional defects (Figure 5). Vector copy numbers per cell in spleen, bone marrow, and blood of transplanted mice ranged from 0.1 to 15, showing efficient repopulation of recipient mice with transduced AWW and WW Lin  cells (Figure S9A). Both LVs were equally efficient in restoring normal monocyte counts in a hemogram (Figure S9). More importantly, transplanted mice achieved equivalent ratios of B220 + ,Gr1 + , and CD11b + cells (Figure 5A, right) as compared to Figure 2. Immortalized “MK-like” Cell Lines Are Relevant Models to Study Wasp Expression during MK Differentiation (A) Scheme of megakaryocytic differentiation protocol of K562 cells (top) and associated morphological changes after 96 h of PMA incubation. (B) Diagram of MK differentiation protocol of MEG-01 cells (top) and morphological changes after VPA addition. (C) Representative FACS plots of CD41a and CD42b megakaryocytic markers of K562 cells in the absence or presence of PMA. (D) Percentage of WASP + K562 cells and WASP expression without PMA treatment, analyzed in the total population (n = 4). (E) Phenotypic changes of MEG-01 cells after VPA treatment expressing CD41a and CD42b markers. (F) Percentage of WASP + MEG-01 cells and WASP expression at 0, 7, 14, and 21 days of VPA-megakaryocytic differentiation (n = 3). Represented data are means ±SEM, non-parametric Mann-Whitney test, two-tailed. **p < 0.01, ***p < 0.001. Molecular Therapy: Methods & Clinical Development 224 Molecular Therapy: Methods & Clinical Development Vol. 19 December 2020 Figure 3. EGFP-LV Expression Driven by WAS Promoter Mimic Endogenous WAS Protein Expression Kinetics during Megakaryocytic Differentiation (A) Schematic representation of the second-generation LVs used in this study expressing enhanced GFP (EGFP). WE 21 harbors the 500-bp core of the WAS proximal promoter (GenBank: AC115618.3 Seq 121356–121851), while AWE 26 (GenBank: AC115618.3: Seq 114605–114991 linker of 17bp/ Seq 121356–121851) and cAWE (GenBank: AC115618.3: Seq 114403–115000 linker of 17 bp / Seq 121356–121851) contain additional 386-bp (Alt) and 597-bp (complete Alt) fragments of the WAS alternative promoter, respectively. (B) Experimental protocols for the comparison of WASP endogenous levels with EGFP expression given by the different LVs during MK differentiation in the different models. (C) Fold expression of endogenous WASP (black bars) and EGFP (colored bars) in K562 cells transduced with WE, AWE, and cAWE LVs (MOI of 1) are related to the expression levels of CD41  CD42  cells in the PMA + condition (non-parametric Mann-Whitney test, *p < 0.05) (n = 4). (D) WASP and EGFP fold (legend continued on next page) www.moleculartherapy.org Molecular Therapy: Methods & Clinical Development Vol. 19 December 2020 225 WT mice, whereas WASKO mice presented reduced levels of B220 + cells and increased levels of Gr1 + and CD11b + cells. Similarly, the efficacy of both LVs to rescue T cell responses (Figures 5B and 5C) and the PLT counts (Figure 5D) were equivalent. These experiments validate the efficacy of both WAS promoter-driven LVs as tools for WAS GT, but they cannot differentiate any superiority. Third-Generation WAS-Driven LVs Have Improved Physiological Expression and Restore Functional Defects In Vitro and In Vivo We finally generated third-generation LVs based on the AWW and investigated their behavior during MK differentiation as well as their therapeutic activity in WASKO mice. We used the WW1.6 LV 15 (Figure 6A, top; this is the LV used in ongoing clinical trials for WAS GT) as the backbone to construct the third-generation AW LV by replacing the full-length (1.6 kb) WAS proximal promoter by the chimeric WAS promoter (0.88 kb) containing sequences from the proximal and alternative promoter (Figures 6AandS2). WAS patient HSPCs were transduced with WW1.6 and AW LVs at an MOI of 50 to achieve similar transduction efficacies (WW1.6 = 2.6 ±1.1 vcn and AW = 3.1 ±1.6 vcn and Figure S5F). Transduced cells were differentiated into MKs and the different subpopulations were analyzed for WAS expression (Figures 6BandS9). As in second-generation LVs, the WASP expression dropped during MK differentiation in both LVs (Figure 6C). We could not find any differences between both LVs in either terms of expression kinetics (Figures 6C) or in PLT expression levels (Figure 6D). However, a comparison of the WASP expression levels of each vector with WASP expression in PLTs from HDs showed significant differences with PLTs derived from 1.6WW-transduced HPSCs, but not with PLTs derived from AW-transduced HPSCs, although more experiments are necessary to demonstrate this. We next evaluated the functional restoration of MKs and PLTS derived from WW1.6and AW-transduced WAS HSPCs by measuring PAC-1 expression in response to thrombin (Figures 6E), the formation of actin nodules (Figures 6F and 6G), or calcium uptake (Figure S11). We found a very similar improvement with both LVs, restoring the response of MKs and PLTs to thrombin, as well as the ability of PLTs to form actin nodules per PLT. Similar findings were also observed for the restoration of podosome formation and clustering on macrophages derived from WAS HSPCs transduced with WW1.6 and AW LVs (Figures 6H– 6K), which exhibited similar WASP expression in the CD33 + CD14 + cells obtained in vitro (Figure 6I). We finally analyzed the therapeutic efficacy of WW1.6 and AW LVs in the WASKO mouse model. WASKO mHSPCs were isolated from bone marrow, transduced with both LVs, and transplanted into irradiated WASKO mice (Figure 6l). Mock-transduced WASKO mHSPCs and WT mHSPCs were injected into control mice. 7 months later, treated mice were sacrificed to analyze the ability of WW1.6 and 0.88-kb AW (AW0.88kb) to restore functional defects. Both LVs were equally efficient in restoring spleen size (Figure 6M) and PLT counts (Figure 6N). DISCUSSION GT has demonstrated to be a real therapeutic alternative for WAS patients. 7,11–14 The current GT approaches for WAS used autologous HSPCs (CD34 + ) genetically modified using a 1.6-kb WAS-proximal promoter-driven LV and transplanted back into the patients. The clinical efficacy of these LVs has been clearly demonstrated since treated patients have shown a reduction of infections and diminished severity of eczema as well as reduced frequency and severity of bleeding episodes. However, although the bleeding was reduced or completely corrected in most patients, 7,14 microthrombocytopenia persisted and normalized PLT counts were not achieved in most patients. 11,14,16 The reasons behind the low PLT recovery achieved with GT compared with the other immune functions are unclear. It was postulated that PLT recovery could correlate with the number of transduced HSPCs reinfused. In this case, the absence of a potent selective advantage in MKs and MK progenitors expressing WASP could be a contributory factor in contrast to lymphocytic lineages. 17 Other hypotheses are that the WW1.6 LV achieved suboptimal levels of WASP expression per copy of integrated vector in MKs, MK progenitors, and/or PLTs. 14 At the same level of HPC transduction, it may be possible to correct the function of lymphoid cells but not that of MK lineage cells. Furthermore, PLTs with suboptimal WASP levels could undergo accelerated destruction in the spleen and in other organs. An ideal GT vector for WAS should not only transduce HSPCs efficiently,butalsomimicendogenousWASPexpressionduringHSPCdifferentiation to the different hematopoietic lineages. Our hypothesis is that physiological expression of WASP during MK differentiation and in PLTs would improve GT outcomes. The aim of this work was therefore to investigate WASP expression along MK differentiation and to definewhichtherapeuticLVmimicsmorecloselyrepresentthispattern. Although WASP expression levels are well established in most mature blood cells, 30,31 the WASP expression kinetics during MK differentiation and thrombopoiesis are not largely studied. We used immortalized cellular models (K562 and MEG-01) as well as primary HSPCs to investigate WASP expression kinetics during these processes. Our data showed an initial decrease in WASP expression from HSPCs to MK progenitors, a slight increase when they are differentiated into early MKs, and a gradual downregulation as these MKs mature into PLTs. This WASP expression kinetic has never been described before and it could be of relevance for normal MK differentiation and thrombopoiesis. We corroborated this expression pattern on two immortalized cell lines, K562 and MEG-01, that resemble different stages of MK differentiation. K562 cells have been described previously as a multipotent myeloid-MK cell line for the study of early events in MK differentiation, 35,36 while MEG-01 cells are defined as an immature MK cell line able to differentiate into mature MKs and PLTs. 37,38 Interestingly, upon MK differentiation, K562 cells mimicked the differentiation expression at days 7 and 14 of MK differentiation of MEG-01 cells treated with VPA (non-parametric Mann-Whitney test, *p < 0.05, **p < 0.01) related to undifferentiated cells (n = 4). (E) WASP and EGFP fold expression in human CD34 + , CD41 + SSC low , CD41 + SSC high , and CD41 + CD42 + cells derived from HSPCs (endogenous WASP, black bars; EGFP, colored bars) at 4, 8, 11, and 16 days after in vitro MK differentiation (n = 10). Non-parametric Mann-Whitney test, two-tailed ( # p < 0.05, ## p < 0.01, ### p < 0.001) and two-way ANOVA, Bonferroni post-test (***p < 0.001) compared to wild-type (WT). All data are represented as mean ±SEM. Molecular Therapy: Methods & Clinical Development 226 Molecular Therapy: Methods & Clinical Development Vol. 19 December 2020 Figure 4. WAS Promoter-Driven LVs Harboring Sequences from the Alternative Promoter Rescue WASP Expression Kinetics in Human Cellular Models of WAS (A) Schematic representation of second-generation LVs expressing WAS cDNA. WW 21 harbors the 500-bp core of the WAS proximal promoter, AWW, 26 and cAWW (the present study). LVs contain an additional 386and 600-bp fragment of the WAS alternative promoter, respectively (see Figure S2 for details). (B) Experimental diagram of K562 WASKO transduced with WAS LVs (MOI of 1) and differentiated to MK-like cells with PMA stimulation. (C and D) Graphs showing the percentage of WASP + cells (C) and WASP expression levels (D) in WT K562 cells (black bars) and in WASKO K562 cells transduced with WW (gray bars), AWW (blue bars), and cAWW (green bars) after MK differentiation with PMA in CD41  CD42  and CD41 + populations (n = 7). (E) Graph shows relative WASP expression in HD CD34 + cells and WAS patient cells transduced with WW (gray), AWW (blue), and cAWW (green) (MOI of 50) at 4, 8, 11, and 16 days of MK in vitro differentiation (n = 4). ***p < 0.001, two-way ANOVA, Bonferroni post-test (compared to same day and conditions of HD). (F) Graph showing WASP expression levels of PLTs derived from WAS patient’s HSPCs transduced with WW, AWW, and cAWW related to WT-driven WASP expression. (G) Representative confocal images of actin nodule formation of platelets derived from HD CD34 + cells (left), WAS patient cells (second left), and WAS patient cells transduced with WW (middle), AWW (second right), and cAWW (right). Several nodules are indicated with white arrows. (H) Quantification of actin nodule per PTL counted in images from (G). Photographs analyzed: HD, 31; WAS, 6; WW, 17; AWW, 26; cAWW, 22. Non-parametric Mann-Whitney test, two-tailed. www.moleculartherapy.org Molecular Therapy: Methods & Clinical Development Vol. 19 December 2020 227 phase from MK progenitors into early MKs, increasing WASP levels, while MEG-01 cells mimic the differentiation stage from early MKs into mature MKs and PLTs, reducing WASP expression. The exact role of these changes in WASP expression during MK differentiation is not clear; however, because the absence of WASP increases PLT production with abnormal phenotype, 18–20 it could be relevant for GT strategies to mimic it. Therefore, our hypothesis is that WAS GT strategies should not only achieve good WASP expression levels on PLTs and MK differentiated cells but also mimic the endogenous expression kinetic during MK differentiation. Once we had established the physiological WASP expression pattern during megakaryopoiesis and thrombopoiesis, we analyzed the behavior of different LVs harboring different fragments of the WAS proximal and alternative promoters. Our previous data indicated that an LV driving EGFP through a WAS promoter (AWE), containing a 386-bp fragment of the alternative promoter immediately upstream of the 500-bp WAS-proximal promoter, increased GFP expression in myeloid cells and mature MKs. 26 The alternative promoter is located 6 kb upstream of the proximal promoter and contains several transcription factor binding sites (TFBSs) shared with the proximal promoter (Sp-1, AP-2, c-Myb, and EGR2) and others that are unique (C/EBP, CP1, Ets-2, and GCF). 28,39 The differences found between both promoters suggest that their activity may vary depending on the cell lineage and the state of differentiation or development. In particular, the presence of TFBSs for C/EBP, CP1, c-Myb, Ets-2, and PU.1 suggest that the alternative promoter could play important roles in the myeloid and megakaryocytic lineages. 40–42 However, the 386-bp fragment of the alternative promoter does not include binding sites for C/EBP and CP1, which are required for higher expression levels in K562 cells. 39 Based on these data, we generated new WAS-driven LVs (cAWE and cAWW) harboring the complete alternative promoter and the 500-bp proximal promoter and compared their behavior with the WAS-proximal promoter LVs (WE and WW), the shortest version of the alternative promoter (386 bp), and the 500-bp proximal promoter (AWE and AWW). The different second-generation LV backbones (W-, AW-, and cAW-) were studied based on their ability to express EGFP and WAS cDNA during MK differentiation and analyzing whether this expression followed the same kinetic as that of the endogenous Figure 5. WWand AWW-Transduced WASKO mHSPCs Rescue Phenotypic Defects of WASKO Mice (A) Restoration of the hematopoietic profile in mice transplanted with WWand AWW-transduced WASKO mHSPCs (MOI of 100; WW mHSPCs = 8.2 vcn/c and AWW = 9.8 vcn/c). Side and forward scatter (left) of cells from PB of the transplanted mice 2 (left, up) and 6 (left, down) months after transplant. The non-transduced (NT) transplanted mice showed a higher number of granulocytes and a lower number of lymphocytes; meanwhile, the WT mHSPC and WASKO mHSPC AWWand WW-corrected mice showed a normal distribution of the population in PB. 6 months post-transplant (right graph), mice were sacrificed and PB was analyzed for vcn/v (see Figure S9), CD45.2 (donor cells), and lineage markers CD3 (T cells), CD220 (B cells), Gr1 + (granulocytes), and CD11b (myeloid cells). Mice per group: WT = 2, NT = 2, AWW = 3, WW0.5 = 3. *p < 0.05, by unpaired t test. Data represent mean ±SEM. (B–D) Phenotypic rescue of WASKO mice after transplantation with WWand AWW-transduced WASKO mHSPCs (MOI of 100; WW mHSPCs = 8 vcn/c and AWW = 6 vcn/c). (B and C) Mice were sacrificed after 3 months, and T cell proliferation assay (CFSE) (B) and IL-2 production (C) were measured after 5 or 3 days, respectively, after stimulation with anti-CD3 (1 mg/mL) by ELISA. One-tailed Mann-Whitney test. Data represent mean ±SEM (WT, N = 2; NT/KO, N = 2; AWW, N = 5; WW, N = 4). *p < 0.05, **p < 0.01. (D) Number of platelets in blood of transplanted mice compared with WT and NT (WT, N = 2, NT, N = 1, AWWN= 4, WW0.5 = 3). Copies/cell (cc) in spleen: AWW, 1.9 ±1.1 cc; WW0.5, 1.8 ±0.4 cc. Data represent mean ±SEM. Molecular Therapy: Methods & Clinical Development 228 Molecular Therapy: Methods & Clinical Development Vol. 19 December 2020 hematopoietic cells by WASP gene proximal promoter sequences. Gene Ther. 12, 715–723. 22. Dupré, L., Marangoni, F., Scaramuzza, S., Trifari, S., Hernández, R.J., Aiuti, A., Naldini, L., and Roncarolo, M.G. (2006). Efficacy of gene therapy for WiskottAldrich syndrome using a WAS promoter/cDNA-containing lentiviral vector and nonlethal irradiation. Hum. Gene Ther. 17, 303–313. 23. Bosticardo, M., Draghici, E., Schena, F., Sauer, A.V., Fontana, E., Castiello, M.C., Catucci, M., Locci, M., Naldini, L., Aiuti, A., et al. (2011). Lentiviral-mediated gene therapy leads to improvement of B-cell functionality in a murine model of Wiskott-Aldrich syndrome. J. Allergy Clin. Immunol 127, 1376–1384.e5. 24. Toscano, M.G., Frecha, C., Benabdellah, K., Cobo, M., Blundell, M., Thrasher, A.J., García-Olivares, E., Molina, I.J., and Martin, F. (2008). Hematopoietic-specific lentiviral vectors circumvent cellular toxicity due to ectopic expression of Wiskott-Aldrich syndrome protein. Hum. Gene Ther. 19,179–197. 25. Cesana, D., Ranzani, M., Volpin, M., Bartholomae, C., Duros, C., Artus, A., Merella, S., Benedicenti, F., Sergi Sergi, L., Sanvito, F., et al. (2014). Uncovering and dissecting the genotoxicity of self-inactivating lentiviral vectors in vivo. Mol. Ther. 22,774–785. 26. Frecha, C., Toscano, M.G., Costa, C., Saez-Lara, M.J., Cosset, F.L., Verhoeyen, E., and Martin, F. (2008). Improved lentiviral vectors for Wiskott-Aldrich syndrome gene therapy mimic endogenous expression profiles throughout haematopoiesis. Gene Ther. 15, 930–941. 27. Petrella, A., Doti, I., Agosti, V., Giarrusso, P.C., Vitale, D., Bond, H.M., Cuomo, C., Tassone, P., Franco, B., Ballabio, A., et al. (1998). A 50regulatory sequence containing two Ets motifs controls the expression of the Wiskott-Aldrich syndrome protein (WASP) gene in human hematopoietic cells. Blood 91, 4554–4560. 28. Hagemann, T.L., and Kwan, S.P. (1999). The identification and characterization of two promoters and the complete genomic sequence for the Wiskott-Aldrich syndrome gene. Biochem. Biophys. Res. Commun. 256, 104–109. 29. Toscano, M.G., Benabdellah, K., Muñoz, P., Frecha, C., Cobo, M., and Martín, F. (2009). Was cDNA sequences modulate transgene expression of was promoterdriven lentiviral vectors. Hum. Gene Ther. 20,1279–1290. 30. Stewart, D.M., Treiber-Held, S., Kurman, C.C., Facchetti, F., Notarangelo, L.D., and Nelson, D.L. (1996). Studies of the expression of the Wiskott-Aldrich syndrome protein. J. Clin. Invest. 97, 2627–2634. 31. Parolini, O., Berardelli, S., Riedl, E., Bello-Fernandez, C., Strobl, H., Majdic, O., and Knapp, W. (1997). Expression of Wiskott-Aldrich syndrome protein (WASP) gene during hematopoietic differentiation. Blood 90,70–75. 32. Dhenge, A., Kuhikar, R., Kale, V., and Limaye, L. (2019). Regulation of differentiation of MEG01 to megakaryocytes and platelet-like particles by valproic acid through Notch3 mediated actin polymerization. Platelets 30, 780–795. 33. Poulter, N.S., Pollitt, A.Y., Davies, A., Malinova, D., Nash, G.B., Hannon, M.J., Pikramenou, Z., Rappoport, J.Z., Hartwig, J.H., Owen, D.M., et al. (2015). Platelet actin nodules are podosome-like structures dependent on Wiskott-Aldrich syndrome protein and ARP2/3 complex. Nat. Commun. 6, 7254. 34. Snapper, S.B., Rosen, F.S., Mizoguchi, E., Cohen, P., Khan, W., Liu, C.H., Hagemann, T.L., Kwan, S.P., Ferrini, R., Davidson, L., et al. (1998). Wiskott-Aldrich syndrome protein-deficient mice reveal a role for WASP in T but not B cell activation. Immunity 9,81–91. 35. Gewirtz, A.M., Burger, D., Rado, T.A., Benz, E.J., Jr., and Hoffman, R. (1982). Constitutive expression of platelet glycoproteins by the human leukemia cell line K562. Blood 60, 785–789. 36. Tabilio, A., Pelicci, P.G., Vinci, G., Mannoni, P., Civin, C.I., Vainchenker, W., Testa, U., Lipinski, M., Rochant, H., and Breton-Gorius, J. (1983). Myeloid and megakaryocytic properties of K-562 cell lines. Cancer Res. 43, 4569–4574. 37. Ogura, M., Morishima, Y., Ohno, R., Kato, Y., Hirabayashi, N., Nagura, H., and Saito, H. (1985). Establishment of a novel human megakaryoblastic leukemia cell line, MEG-01, with positive Philadelphia chromosome. Blood 66, 1384–1392. 38. Takeuchi, K., Ogura, M., Saito, H., Satoh, M., and Takeuchi, M. (1991). Production of platelet-like particles by a human megakaryoblastic leukemia cell line (MEG-01). Exp. Cell Res. 193, 223–226. 39. Hagemann, T.L., Mares, D., and Kwan, S. (2000). Gene regulation of Wiskott-Aldrich syndrome protein and the human homolog of the Drosophila Su(var)3-9: WASP and SUV39H1, two adjacent genes at Xp11.23. Biochim. Biophys. Acta 1493, 368–372. 40. García, P., Berlanga, O., Vegiopoulos, A., Vyas, P., and Frampton, J. (2011). c-Myb and GATA-1 alternate dominant roles during megakaryocyte differentiation. J. Thromb. Haemost. 9, 1572–1581. 41. Postigo, A.A., Sheppard, A.M., Mucenski, M.L., and Dean, D.C. (1997). c-Myb and Ets proteins synergize to overcome transcriptional repression by ZEB. EMBO J. 16, 3924–3934. 42. Reddy, M.A., Yang, B.S., Yue, X., Barnett, C.J., Ross, I.L., Sweet, M.J., Hume, D.A., and Ostrowski, M.C. (1994). Opposing actions of c-ets/PU.1 and c-myb protooncogene products in regulating the macrophage-specific promoters of the human and mouse colony-stimulating factor-1 receptor (c-fms) genes. J. Exp. Med. 180, 2309–2319. 43. Zufferey, R., Dull, T., Mandel, R.J., Bukovsky, A., Quiroz, D., Naldini, L., and Trono, D. (1998). Self-inactivating lentivirus vector for safe and efficient in vivo gene delivery. J. Virol. 72, 9873–9880. 44. Benabdellah, K., Muñoz, P., Cobo, M., Gutierrez-Guerrero, A., Sánchez-Hernández, S., Garcia-Perez, A., Anderson, P., Carrillo-Gálvez, A.B., Toscano, M.G., and Martin, F. (2016). Lent-On-Plus lentiviral vectors for conditional expression in human stem cells. Sci. Rep. 6, 37289. 45. Charrier, S., Stockholm, D., Seye, K., Opolon, P., Taveau, M., Gross, D.A., BucherLaurent, S., Delenda, C., Vainchenker, W., Danos, O., and Galy, A. (2005). A lentiviral vector encoding the human Wiskott-Aldrich syndrome protein corrects immune and cytoskeletal defects in WASP knockout mice. Gene Ther. 12, 597–606. www.moleculartherapy.org Molecular Therapy: Methods & Clinical Development Vol. 19 December 2020 235