Tissue-resident myeloid (TRM) cells in adults have highly variable lifespans, and may be derived from early embryonic yolk sac, fetal liver, or bone marrow. Some of these TRM cells are known pathogenic participants in congenital and acquired diseases. Myeloablative conditioning and hematopoietic stem cell transplantation can replace long-lived brain TRM cells, resulting in clinical improvements in metabolic storage diseases. With the advent of antibody-drug conjugate (ADC)-targeted cell killing as a cell-selective means of transplant conditioning, we assessed the impact of anti-CD45-ADC on TRM cells in multiple tissues. Replacement of TRM cells ranged from 40% to 95% efficiencies in liver, lung, and skin tissues, after a single anti-CD45-ADC dose and bone marrow hematopoietic cell transfer. Of note, the population size of TRM cells in tissues returned to pretreatment levels, suggesting a regulated control of TRM cell abundance. As expected, brain microglia were not affected, but brain monocytes and macrophages were 50% replaced. Anti-CD45-ADC and adoptive cell transfer were then tested in the chronic acquired condition, atherosclerosis exacerbated by Tet2 mutant clonal hematopoiesis. Plaque-resident myeloid cells were efficiently replaced with anti-CD45-ADC and wild-type bone marrow cells. Notably, this reduced existent atherosclerotic plaque burden. Overall, these results indicate that the anti-CD45-ADC clears both hematopoietic stem and TRM cells from their niches, enabling cell replacement to achieve disease modification in a resident myeloid cell-driven disease.

Tissue resident myeloid cells (TRM) in adults have highly variable lifespans and may be derived from early embryonic yolk sac, fetal liver or bone marrow. Some of these TRM are known pathogenic participants in congenital and acquired diseases. Myeloablative conditioning and hematopoietic stem cell transplant can replace long-lived brain TRM resulting in clinical improvements in metabolic storage diseases. With the advent of antibody-drug-conjugate (ADC) targeted cell killing as a cell selective means of transplant conditioning, we assessed the impact of anti-CD45-ADC on TRM in multiple tissues. Replacement of TRM ranged from 40 to 95 percent efficiencies in liver, lung, and skin tissues, after a single anti-CD45-ADC dose and bone marrow hematopoietic cell transfer. Of note, the population size of TRM in tissues returned to pre-treatment levels suggesting a regulated control of TRM abundance. As expected, brain, microglia were not affected, but brain monocytes and macrophages were 50% replaced. Anti-CD45-ADC and adoptive cell transfer were then tested in the chronic acquired condition, atherosclerosis exacerbated by Tet2 mutant clonal hematopoiesis. Plaque resident myeloid cells were efficiently replaced with anti-CD45-ADC and wild-type bone marrow cells. Notably, this reduced existent atherosclerotic plaque burden. Overall, these results indicate that anti-CD45-ADC clears both HSC and TRM niches enabling cell replacement to achieve disease modification in a resident myeloid cell driven disease.

Autologous hematopoietic stem cell transplantation (Auto-HSCT) with gene-modification techniques represents a potential cure for multiple genetic blood diseases. Despite its broad curative potential, auto-gene modified HSCT is currently limited due to morbidity/mortality from cytotoxic chemotherapy-based conditioning, including risks of secondary malignancies, organ toxicity, and infertility. To overcome these limitations, we have developed antibody drug conjugates (ADC) targeting CD117 (C-KIT) to specifically deplete the hematopoietic stem and progenitor cells (HSPC) prior to auto-gene modified HSCT. We have previously shown that the anti-CD117 ADC is highly effective at killing human CD117+ cells in vitro and in vivo (Pearse et al., Blood 2018 132:3314). To validate CD117 as an appropriate antigen for targeted ADC-mediated depletion prior to HSCT, we developed an optimized non-human primate (NHP) tool anti-CD117 ADC and evaluated it in an auto-gene modified HSCT in the rhesus macaque model. The tool CD117-ADC is potent on primary human and NHP CD34+ cells in vitro with EC50 of 0.2 and 0.09 pM respectively (Figure 1A). Humanized NSG mice treated with the tool CD117-ADC had full depletion of human HSPCs in the bone marrow 21 days after a single administration of the ADC, while maintaining the peripheral immune cells. We next tested the efficacy and safety of the tool CD117-ADC in NHPs. A single administration of the tool CD117-ADC was fully myeloablative (>99% HSPC depletion) and comparable to HSPC depletion observed following busulfan conditioning (6 mg/kg, once daily for 4 consecutive days). There was no effect on the peripheral and bone marrow lymphocytes and the ADC was well tolerated. To facilitate the use in HSCT, the tool CD117-ADC was engineered to have a fast clearance and in this study the half-life was <10 hours. Based on these encouraging results, we explored whether the tool CD117-ADC could enable engraftment of autologous gene modified hematopoietic stem cells in the rhesus macaque model. A single rhesus macaque was mobilized with granulocyte-colony stimulating factor (G-CSF, 20 mcg/kg/day x 5) and plerixafor (1 mg/kg on day 5 of G-CSF) prior to apheresis. The isolated CD34+ cells were transduced with a lentivirus encoding the β-globin gene and cryopreserved. The tool CD117-ADC was dosed on day -6 and the cryopreserved gene modified cells were thawed and infused (3.3 x 106 CD34+ cells/kg) on day 0. A bone marrow aspirate analyzed on the day of infusion (day 0) demonstrated >99% depletion of the HSPCs and preserved of the bone marrow lymphocytes (Figure 1B). The primate engrafted neutrophils and platelets on day 8 and 10 respectively, and the peripheral lymphocytes were maintained throughout the transplant (Figure 1C). The gene marking in the granulocytes was detectable at day 9, and additional follow up and data from additional animals will be presented. In summary, we have developed a tool CD117 ADC that shows potent activity on NHP CD34+ cells. This optimized CD117-ADC is fully myeloablative with a single dose in NHPs, has a favorable safety profile, spares the immune system and is cleared rapidly as designed. In a rhesus model of autologous gene modified HSCT, a single dose of the ADC enables engraftment of auto-gene modified HSC. These proof of concept studies validate the use of CD117-ADC for targeted HSPC depletion prior to transplant and support its use as a new conditioning agent for autologous gene modified HSCT. This targeted approach for safer conditioning could improve the risk benefit profile for patients undergoing stem cell transplant and enable more patients to benefit from these potentially curative therapies. Disclosures Pearse: Magenta Therapeutics: Employment, Equity Ownership, Patents & Royalties. McDonough:Magenta Therapeutics: Employment, Equity Ownership, Patents & Royalties. Proctor:Magenta Therapeutics: Employment, Equity Ownership. Panwar:Magenta Therapeutics: Employment, Equity Ownership. Sarma:Magenta Therapeutics: Employment, Equity Ownership. Kien:Magenta Therapeutics: Employment, Equity Ownership. Latimer:Magenta Therapeutics: Employment, Equity Ownership. Dushime:Magenta Therapeutics: Employment, Equity Ownership. Hyzy:Magenta Therapeutics: Employment, Equity Ownership. Brooks:Magenta Therapeutics: Employment, Equity Ownership. Palchaudhuri:Magenta Therapeutics: Employment, Equity Ownership. Li:Magenta Therapeutics: Employment, Equity Ownership. Sawant:Magenta Therapeutics: Employment, Equity Ownership. McDonagh:Magenta Therapeutics: Employment. Boitano:Magenta Therapeutics: Employment, Equity Ownership, Patents & Royalties. Cooke:Magenta Therapeutics: Employment, Equity Ownership, Patents & Royalties.

Background Allogeneic hematopoietic stem cell transplant (HSCT) is a curative approach to halt disease progression of select inherited metabolic disorders (IMDs). For IMDs that affect the central nervous system (CNS), donor-derived cells, like microglia, cross-correct defects via production of normal enzyme. The typical cell dose used in HSCT can be sub-optimal and, in gene therapy applications, copy number has been shown to be variable. We developed MGTA-456, a high dose cell therapy that led to rapid neutrophil recovery and 100% engraftment in patients with malignant and non-malignant diseases (Wagner et al Blood 2017; Orchard et al AAN 2019). The impact of cell dose on disease correction, however, is unknown. Here, we show that fast and robust hematopoietic and microglia recovery via high cell dose therapies, like MGTA-456, leads to rapid and complete disease resolution of Hurler syndrome through hematopoietic engraftment in the CNS. Results At 1-16 weeks post-HSCT, transplant of increasing doses of total bone marrow cells following myeloablative busulfan conditioning led to a dose-dependent increase microglial engraftment (26-fold for 10 × 106 cells vs 0.3 × 106 cells, p<0.01), with 91-99% peripheral donor chimerism. In Idua−/− mice (Figure A), high cell doses led to an improvement in CNS disease endpoints, such as IDUA enzyme delivery (p<0.05), reduced substrate accumulation, and normalization of behavioral activity to wild type levels at 1 month post-HSCT (Figure B-C). Reduction of substrate to wild type levels was observed as early as 1 week after transplant with 10 × 106 cells (p<0.01). In contrast, transplant of 10 × 106 cells into mice conditioned with a myeloablative dose of treosulfan, an agent that does not permit brain microglial engraftment (Figure A), did not correct CNS defects (Figure B-C), suggesting that donor engraftment in the CNS is required for disease correction.Relative to low cell dose therapies, transplant of MGTA-456, a high cell dose therapy with two normal IDUA gene copies, led to robust, long-term immune recovery (n=88 animals), 60-fold greater microglial engraftment as early as 2 weeks post-HSCT (Figure D, p<0.001), and >600-fold higher IDUA enzyme levels (Figure E, p<0.001). MGTA-456 enabled use of low-dose busulfan, with 21-fold greater microglial engraftment than that achieved by standard approaches using high-dose busulfan (p<0.01, n=8). Mechanistically, brain microglia are derived from CD34+CD90+ cells, which are present at high numbers in MGTA-456. Conclusions We show that high dose HSCT leads to improved disease correction, including normalization of behavioral outcomes, via robust engraftment. High dose cell therapies, like MGTA-456, may rapidly and durably resolve peripheral and neurologic disease in patients with IMDs and other neurodegenerative diseases caused by defective microglia.

For patients with refractory or high-risk hematologic malignancies, like acute myeloid leukemia (AML), myelodysplastic syndromes (MDS), and acute lymphoblastic leukemia (ALL), allogeneic hematopoietic stem cell transplant (Allo-HSCT) is a potentially curative approach. Morbidities and mortality associated with current conditioning regimens limit the use of this curative procedure. As a result, many eligible patients do not consider transplant and 2/3 of those transplanted are only able to tolerate a reduced intensity conditioning regimen, which is associated with increased relapse rates (Scott, J Clin Onc 2017). Thus, there is an urgent need for a safer and more effective conditioning agent with improved disease control. Our targeted antibody drug conjugate (ADC) approach is designed to improve the safety of current conditioning protocols by specifically depleting CD45+ cells. We developed a novel anti-human CD45-targeted short half-life ADC conjugated to amanitin (AM). CD45 is the ideal target for allo-HSCT conditioning because it is expressed on all hematopoietic cells (except erythrocytes, plasma cells and platelets), and most hematologic malignancies. Given their targeted specificity, anti-CD45-AM can potentially provide dual benefit to leukemia patients by combining effective conditioning for HSCT with depletion of target-bearing tumor cells.To demonstrate our engineered short half-life anti-human CD45-AM has anti-leukemic activity we tested it in human leukemic xenograft murine models. A panel of models were evaluated to mimic untreated and refractory disease; AML PDX models (from treatment naïve and relapsed post allogeneic HCT patients), ALL cells from an immortalized cell line (REH-Luc), T-ALL patient-derived xenograft (PDX) model (from a patient progressing post DHAP chemotherapy).In the REH-Luc model, single doses of anti-CD45-AM were well tolerated, and cytoreductive resulting in delayed tumor growth compared to vehicle (PBS), isotype-AM, or standard of care (SoC) doxorubicin. Anti-CD45-AM treatment in the PDX AML, and T-ALL significantly decreased peripheral tumor burden resulting in delayed tumor growth compared to vehicle, isotype-AM, and comparable to 2 clinically validated standards of care (Ara-C, and dexamethasone respectively; figure 1). As designed for the transplant indication, the ADC had a reduced half-life compared to wild type antibody controls (16 vs 79h).These data in humanized murine xenograft models demonstrate that short half-life CD45-AM ADCs are potent targeted anti-leukemia agents. Together with prior reports demonstrating the potency of anti-CD45-AM as conditioning agents, these non-genotoxic ADCs may be useful in reducing disease burden and inducing durable remissions in patients after transplant particularly those who receive reduced intensity conditioning that are at high risk of relapse. For patients with refractory or high-risk hematologic malignancies, like acute myeloid leukemia (AML), myelodysplastic syndromes (MDS), and acute lymphoblastic leukemia (ALL), allogeneic hematopoietic stem cell transplant (Allo-HSCT) is a potentially curative approach. Morbidities and mortality associated with current conditioning regimens limit the use of this curative procedure. As a result, many eligible patients do not consider transplant and 2/3 of those transplanted are only able to tolerate a reduced intensity conditioning regimen, which is associated with increased relapse rates (Scott, J Clin Onc 2017). Thus, there is an urgent need for a safer and more effective conditioning agent with improved disease control. Our targeted antibody drug conjugate (ADC) approach is designed to improve the safety of current conditioning protocols by specifically depleting CD45+ cells. We developed a novel anti-human CD45-targeted short half-life ADC conjugated to amanitin (AM). CD45 is the ideal target for allo-HSCT conditioning because it is expressed on all hematopoietic cells (except erythrocytes, plasma cells and platelets), and most hematologic malignancies. Given their targeted specificity, anti-CD45-AM can potentially provide dual benefit to leukemia patients by combining effective conditioning for HSCT with depletion of target-bearing tumor cells. To demonstrate our engineered short half-life anti-human CD45-AM has anti-leukemic activity we tested it in human leukemic xenograft murine models. A panel of models were evaluated to mimic untreated and refractory disease; AML PDX models (from treatment naïve and relapsed post allogeneic HCT patients), ALL cells from an immortalized cell line (REH-Luc), T-ALL patient-derived xenograft (PDX) model (from a patient progressing post DHAP chemotherapy). In the REH-Luc model, single doses of anti-CD45-AM were well tolerated, and cytoreductive resulting in delayed tumor growth compared to vehicle (PBS), isotype-AM, or standard of care (SoC) doxorubicin. Anti-CD45-AM treatment in the PDX AML, and T-ALL significantly decreased peripheral tumor burden resulting in delayed tumor growth compared to vehicle, isotype-AM, and comparable to 2 clinically validated standards of care (Ara-C, and dexamethasone respectively; figure 1). As designed for the transplant indication, the ADC had a reduced half-life compared to wild type antibody controls (16 vs 79h). These data in humanized murine xenograft models demonstrate that short half-life CD45-AM ADCs are potent targeted anti-leukemia agents. Together with prior reports demonstrating the potency of anti-CD45-AM as conditioning agents, these non-genotoxic ADCs may be useful in reducing disease burden and inducing durable remissions in patients after transplant particularly those who receive reduced intensity conditioning that are at high risk of relapse. Figure 1.

Hematopoietic stem cell transplant (HSCT) can be a highly effective, and often curative, treatment for patients with AML. At present, myeloablative conditioning (MAC) regimens are associated with severe acute and long-term toxicities. A subset of transplant-eligible AML patients undergoes reduced-intensity conditioning (RIC) regimens; while reducing toxicities, RIC regimens have higher incidence of post-transplant relapse and graft failure vs MAC. Thus, the need for safe and effective targeted conditioning agents for a broader use in transplant in both malignant as well as non-malignant settings is critical. We developed MGTA-117, an anti-human CD117 (c-kit)-targeted antibody (Ab) engineered for a short-half life (t1/2) conjugated to amanitin, an RNA polymerase II inhibitor, to enable ADC clearance prior to HSCT. MGTA-117 has a t1/2 of 91 hours in humanized NSG (hNSG) mice compared to the parental wild type (WT) Ab with a t1/2 of 217 hours in human FCRN mice. Via optimization of the linker-toxin moiety, the maximum tolerated dose (MTD) in C57BL/6 mice was improved 17-fold compared to previous generations of the ADC. The antibody component specifically binds CD117, which is expressed on hematopoietic stem and progenitor cells (HSPC) and in ~80% of patient cells with AML and MDS (Gao et al. PLOS One. 2015). When conjugated to amanitin, this ADC robustly depletes both CD117+ HSPCs and leukemic blasts. This targeted and optimized approach not only broadens the therapeutic window across preclinical models, but also provides the dual benefit of effective conditioning for HSCT and reduction of target-expressing tumor cells. Previously we have shown MGTA-117 elicits potent cytotoxicity on both primary human CD34+ cells and a CD117+ AML cell line in vitro. In hNSG mice, a single dose (0.1 mg/kg) of MGTA-117 selectively depleted ≥95% human HSPCs (Figures 1A and B). To demonstrate anti-leukemic activity of MGTA-117, we studied it in multiple human leukemic xenograft murine models that mimic untreated and refractory AML, including both cell line derived (Kasumi-1, CD117-expressing leukemia cell line, in NSG mice) and patient derived (AML PDX 1: J000106134, treatment naïve AML PDX and AML PDX 2: J000106132, heavily pretreated relapsed refractory AML PDX in NSG-SGM3 mice) xenograft models. MGTA-117 was well-tolerated in all three AML xenograft models both as a single dose (1-10 mg/kg) and as a multi-dose (3 mg/kg QODx2) regimen. In the Kasumi-1 model (Figure 1C), MGTA-117, both as a single dose (3 & 10 mg/kg) and as a multi-dose (3 mg/kg QODx2), resulted in a 2.4-2.7-fold increase in median survival compared to vehicle (PBS), isotype-ADC, or standard of care (SOC) cytarabine (ARA-C, 30 mg/kg QDx5). Furthermore, a single dose of MGTA-117 significantly decreased peripheral tumor burden leading to delayed tumor growth resulting in a 2-3.3 fold and 1.3-1.8 fold increase in median survival in the treatment naïve AML PDX 1 and highly pretreated AML PDX 2 models (respectively) compared to vehicle, isotype-ADC and SOC. ARA-C treatment demonstrated a modest improvement in survival in both the AML PDX 1 (1.3 fold), AML PDX 2 (1.3 fold) models compared to control. These data demonstrate that MGTA-117 has potential to be a potent targeted conditioning and anti-leukemia agent. Together with prior reports demonstrating MGTA-117 as an effective conditioning agent in an animal model, this targeted ADC approach has the potential to improve HSCT outcomes in AML by reducing leukemic burden peri-transplant and may decrease the toxicities associated with current conditioning regimens. Disclosures Lanieri: Magenta Therapeutics: Current Employment. Lamothe:Magenta Therapeutics: Current Employment. Miske:Magenta Therapeutics: Current Employment. McDonough:Magenta Therapeutics: Current Employment. Sarma:Magenta Therapeutics: Ended employment in the past 24 months. Bhattarai:Magenta Therapeutics: Current Employment. Latimer:Magenta Therapeutics: Current Employment. Dushime:Magenta Therapeutics: Current Employment. Jain:Magenta Therapeutics: Current Employment. Palchaudhuri:Magenta Therapeutics: Current Employment. Knihtila:Magenta Therapeutics: Current Employment. Pearse:Magenta Therapeutics: Ended employment in the past 24 months. Proctor:Magenta Therapeutics: Current Employment. Boitano:Magenta Therapeutics: Ended employment in the past 24 months, Patents & Royalties. Cooke:Magenta Therapeutics: Ended employment in the past 24 months, Patents & Royalties. Davis:Magenta Therapeutics: Current Employment, Current equity holder in publicly-traded company.

Background HSC dose and HLA match are risk factors that impact mortality using cord blood units (CBU) in transplant. Low CD34+ doses result in prolonged cytopenia and higher graft failure risk. A minimum cell dose of 3.0 × 107 total nucleated cells (TNC)/kg has generally been required in CBU selection; however, cell dose limits often require the use of a 2nd unit and markedly limits the availability of 7-8/8 HLA-matched units in larger patients. MGTA-456 is a cell therapy product utilizing an aryl hydrocarbon receptor antagonist for expansion of CD34+ HSCs in vitro. In prior studies with fresh MGTA-456, 36 patients with hematologic malignancies demonstrated rapid neutrophil recovery and 100% engraftment. This study (NCT03674411) will evaluate the safety and efficacy of cryopreserved MGTA-456 and the effectiveness of lowering the TNC threshold of the selected CBU to 1.0 × 107 TNC/kg before expansion to improve HLA match. Methods 13 patients aged 2-47 years (12-159 kg) with high-risk hematologic malignancy were enrolled with 11 transplanted to date. Patients received cyclophosphamide 120 mg/kg, fludarabine 75 mg/m2 and total body irradiation 1320 cGy or a busulfan-based regimen for children ≤3 years of age prior to MGTA-456 with cyclosporine/mycophenolate mofetil prophylaxis. G-CSF began the day after infusion until the neutrophil count exceeded 2500/uL for 3 days. Results MGTA-456 contained a median 2.3 × 109 CD34+ cells (range, 0.65-8.0) after expansion (422-fold expansion of CD34+ cells [range, 219-1313]). Neutrophil recovery occurred in all patients with a median of 15 days (range, 0-31), similar to recipients of fresh MGTA-456 in a prior study (median 14 days) and significantly faster than in recipients of unmodified CBUs (median 25 days). Median platelet recovery was 41 days (range 24-49) vs 64 days with unmodified CBUs. MGTA-456 CD34+CD90+ content strongly correlated with speed of neutrophil recovery (Fig. 1), consistent with preclinical murine data showing CD34+CD90+ cells represent the true engraftable HSC population. Lowering the TNC requirement to 1.0 × 107 TNC/kg for CBU selection pre-expansion improved HLA match and/or eliminated the need for double CBU transplant in all 5 patients weighing >80 kg. Three patients received 8/8 grafts who would have otherwise received 1 or 2 6/8 units (n=2) or two 7/8 units (n=1) which may account for the low incidence of aGVHD overall (3 cases of Grade 1-2 and no Grade 3-4). With a follow-up of 5.2 months (0.4-8.5 months), all patients are alive. Conclusion Cryopreserved MGTA-456 cell therapy resulted in rapid and 100% engraftment with speed of neutrophil recovery correlating with CD34+CD90+ cell dose in patients with high risk hematologic malignancy. Expansion of smaller CBUs increases the chance of finding a better HLA match, particularly for adults, thus reducing the barriers associated with low cell dose and poor HLA match in CBU transplantation.

Introduction Allogeneic hematopoietic stem cell transplant (Allo-HSCT) is a potentially curative treatment for malignant and non-malignant blood disorders. However, current conditioning regimens limit the use of this curative procedure in many eligible patients due to regimen-related mortality and morbidities, including organ toxicity, infertility, and secondary malignancies. We are developing novel antibody drug conjugates (ADC) as conditioning agents that can achieve full myeloablation as a single agent that may reduce toxicity associated with current conditioning regimens. We have generated an anti-murine ADC targeting CD45 and assessed its effectiveness as single agent conditioning regimen in a fully allogeneic murine HSCT model. Methods Our tool CD45 ADC is engineered for rapid clearance (t1/2=1.7hr) to enable HSCT after conditioning. A single dose of 3 mg/kg is fully myeloablative in C57BL/6 mice. To determine if the tool CD45-ADC could successfully condition recipients for fully mismatched allo-HSCT, we evaluated the ability of a single dose of 5 mg/kg of the tool CD45-ADC to condition C57BL/6 hosts (H-2b, CD45.2+) for transplant with cells from CByJ.SJL(B6) donors (H-2d, CD45.1+). A matched dose of an isotype ADC (Iso-ADC) was used as a negative control, while 9 Gy TBI was used as a conventional conditioning positive control. Conditioned mice were transplanted with 4x107 whole BM cells, and peripheral blood chimerism was assessed over 22 weeks. At 22 weeks, donor hematopoietic cell chimerism was evaluated in the spleen, bone marrow, and thymus of recipients. Results In the fully mismatched Balb/c → C57Bl/6 allo-HSCT model, conditioning with a single dose of 5 mg/kg of CD45-ADC as a single agent was well tolerated and enabled full allogeneic donor chimerism (n=2 separate experiments). Peripheral blood chimerism was observed in mice conditioned with CD45-ADC at week 4 and maintained through week 22 (Figure 1). Multilineage reconstitution was observed in the T-, B-, and myeloid cell compartments with >90% donor chimerism seen in each compartment, indicative of HSC engraftment. These results were comparable to chimerism seen in the 9 Gy TBI positive control. Treatment with a non-targeting isotype ADC at a matched dose was not effective (Figure 1). For all groups, stem cell chimerism in the bone marrow matched that in the periphery. Splenic and thymic donor immune cell reconstitution was similar between CD45-ADC and TBI conditioning at week 22 (Figure 1), demonstrating that CD45-ADC efficiently depletes host lymphocytes in secondary lymphoid organs while preserving the capacity of the host thymus to support de novo generation of donor-derived T cells after transplantation. Conclusion Conditioning with CD45-ADC was well-tolerated, fully myeloablative, and enabled complete chimerism in a full mismatch allo-HSCT model as a single agent. This targeted, readily translatable approach for safer conditioning could improve the risk-benefit profile for allogenic and haploidentical HSCT and may extend the curative potential of HSCT to more patients suffering from blood cancers and other diseases that may benefit from HSCT. Disclosures Hyzy: Magenta Therapeutics: Current Employment, Current equity holder in publicly-traded company. Proctor:Magenta Therapeutics: Current Employment. Gillard:Magenta Therapeutics: Current Employment. Hammond:Magenta Therapeutics: Current Employment, Current equity holder in publicly-traded company. Sarma:Magenta Therapeutics: Ended employment in the past 24 months. Clark:Magenta Therapeutics: Current Employment. Bhat:Magenta Therapeutics: Current Employment. Lamothe:Magenta Therapeutics: Current Employment. Palchaudhuri:Magenta Therapeutics: Current Employment. Pearse:Magenta Therapeutics: Ended employment in the past 24 months. McDonagh:Magenta Therapeutics: Ended employment in the past 24 months. Kiem:Magenta Therapeutics: Consultancy; CSL: Consultancy; Homology Medicines: Membership on an entity's Board of Directors or advisory committees; Rocket Pharma: Membership on an entity's Board of Directors or advisory committees; Umoja: Membership on an entity's Board of Directors or advisory committees; Enochian: Membership on an entity's Board of Directors or advisory committees; Vor Biopharma: Membership on an entity's Board of Directors or advisory committees. Wagner:Rocket Pharmaceuticals, Inc.: Consultancy, Current equity holder in publicly-traded company; Novartis: Research Funding; Magenta Therapeutics: Consultancy, Research Funding; BlueRock: Research Funding; Gadeta: Membership on an entity's Board of Directors or advisory committees. Blazar:Magenta Therapeutics: Consultancy; Fate Therapeutics Inc.: Research Funding; BlueRock Therapeutics: Research Funding; Childrens' Cancer Research Fund: Research Funding; KidsFirst Fund: Research Funding; Tmunity: Other: Co-founder; BlueRock Therapeuetic: Consultancy. Boitano:Magenta Therapeutics: Ended employment in the past 24 months, Patents & Royalties. Cooke:Magenta Therapeutics: Ended employment in the past 24 months, Patents & Royalties. Davis:Magenta Therapeutics: Current Employment, Current equity holder in publicly-traded company.

Conditioning chemotherapy is used to deplete hematopoietic stem cells in the recipient's marrow, facilitating donor cell engraftment. Although effective, a major issue with chemotherapy is the systemic genotoxicity that increases the risk for secondary malignancies. Antibody conjugates targeting hematopoietic cells are an emerging non-genotoxic method of opening the marrow niche and promoting engraftment of transplanted cells while maintaining intact marrow cellularity. Specifically, this platform would be useful in diseases associated with DNA damage or cancer predisposition, such as dyskeratosis congenita, Schwachman-Diamond syndrome, and Fanconi anemia (FA). Our approach utilizes antibody-drug conjugates (ADC) as an alternative conditioning regimen in an FA mouse model of autologous transplantation. Antibodies targeting either CD45 or CD117 were conjugated to saporin (SAP), a ribosomal toxin. FANCA knockout mice were conditioned with either CD45-SAP or CD117-SAP prior to receiving whole marrow from a heterozygous healthy donor. Bone marrow and peripheral blood analysis revealed equivalent levels of donor engraftment, with minimal toxicity in ADC-treated groups as compared with cyclophosphamide-treated controls. Our findings suggest ADCs may be an effective conditioning strategy in stem cell transplantation not only for diseases where traditional chemotherapy is not tolerated, but also more broadly for the field of blood and marrow transplantation.

IntroductionAllogeneic hematopoietic stem cell transplant (Allo-HSCT) is a potentially curative treatment for malignant and non-malignant blood disorders. However, current conditioning regimens limit its use due to regimen-related mortality and morbidities. We are developing novel antibody drug conjugates (ADCs) to provide the benefit of full-intensity conditioning to remove disease-causing cells while reducing the severity of treatment-related adverse events. These ADCs are designed to deliver agents specifically to CD45+ target cells as a sole conditioning agent or as the primary conditioning agent in a reduced intensity conditioning protocol for allo-HSCT. The aim of this study was to model this approach with a tool anti-mouse CD45 ADC to determine if targeted ADCs can be used to enable allo-HSCT in mice.MethodsWe developed a tool anti-mouse CD45 ADC engineered to have a short half-life (T1/2 = 1.7hr) to enable HSCT. The optimal dose of tool CD45-ADC was established in a congenic autologous mouse transplant model. Next, the tool CD45-ADC was evaluated alone or in combination with low dose (0.5 Gy) total body irradiation (TBI) or T cell-depleting anti-mouse antibodies (CD4 and CD8, 0.25 mg/kg IP) in a full mismatch allo-HSCT model (Balb/c donors (H-2d, CD45.1+) into C57Bl/6 recipients (H-2b, CD45.2+)). 9 Gy TBI served as the conventional conditioning positive control. Conditioned mice were transplanted with 2 × 107 whole bone marrow cells, and peripheral blood chimerism assessed over 16 weeks.ResultsA single 3 mg/kg dose of the tool CD45-ADC was an effective regimen in a congenic autologous mouse transplant model, resulting in full donor chimerism comparable to conditioning with 9 Gy TBI. In a full mismatch allo-HSCT model, a single dose of CD45-ADC enabled mixed myeloid chimerism out to 3 weeks as a single agent, but the chimerism was transient. In combination with low dose TBI or T cell depleting antibodies, the tool CD45-ADC enabled >90% peripheral donor chimerism by week 4 post-transplantation, which was maintained through week 16. Multilineage reconstitution of T-, B-, and myeloid cell compartments was observed (>90% donor chimerism) and was comparable to chimerism seen in the 9 Gy TBI positive control. Treatment with a non-targeting isotype ADC was not effective (Figure 1A, 1B).ConclusionA single dose of the tool CD45-ADC is fully myeloablative and enables complete chimerism in a full mismatch allo-HSCT model with low dose TBI or supplemental T cell depletion. Future experiments will examine the use of tool CD45-ADC conditioning in HSCT as a treatment in mouse autoimmune disease models. This targeted, readily translatable approach for safer conditioning could improve the risk-benefit profile for allogenic and haploidentical HSCT and may extend the curative potential of this therapeutic modality.

Background . Allogeneic hematopoietic stem cell transplant (HSCT) is a promising approach to halt disease progression and prevent or ameliorate neurological symptoms arising from select inherited metabolic disorders (IMDs). Donor-derived cells, including microglia, limit disease progression post-HSCT via production of normal enzyme in a process called cross-correction. A standard cell dose used in HSCT is sub-optimal, resulting in delayed hematopoietic recovery and slower correction of central nervous system (CNS) defects (Lund et al BBMT 2019). To address these limitations, we developed MGTA-456, a cell therapy that contains large numbers of CD34+ cells and has led to accelerated neutrophil recovery and 100% engraftment post-HSCT in patients with malignant and non-malignant diseases (Wagner et al Blood 2017; Orchard et al AAN 2019). We previously showed that MGTA-456 leads to faster hematopoietic and microglia recovery in the brains of NSG mice (Goncalves et al AAN 2019); however, the impact of cell dose on disease outcomes and mechanism of cross-correction are unknown. Here, we show that faster and greater hematopoietic and microglia recovery leads to rapid and complete resolution of disease endpoints in a mouse model of mucopolysaccharidosis I (Hurler syndrome) and that, mechanistically, donor engraftment in the brain is required for disease cross-correction. Results . To determine whether cell dose impacts microglial engraftment, CD45.2 mice were conditioned with a clinically-relevant, myeloablative dose of busulfan and transplanted with increasing doses of CD45.1 bone marrow cells, beginning with 0.3x106 cells/mouse (2x106 cells/kg) based on allometric scaling to model high dose cell therapies. A dose-dependent increase in microglia was observed as early as 1 week post-HSCT, where 10x106 cells led to a 26-fold higher number of donor microglia compared to 0.3x106 cells (p<0.01), an effect that was sustained through 16 weeks post-HSCT (p<0.001). Despite high donor chimerism in the periphery at all cell doses (75-99%), only partial chimerism was observed in the brain. At 16 weeks, donor microglia represented only 2% of microglia after transplant of 0.3x106 cells but this was increased to 35% of total microglia in the brain following transplant of 10x106 cells. These data indicate that while busulfan can facilitate a low level of microglia engraftment, this effect can be enhanced by transplant of high cell doses. To evaluate the impact of cell dose on disease outcomes, we transplanted a low (0.3x106) or high (10x106) cell dose of wild-type bone marrow cells into busulfan-conditioned Idua-/- mice, a model of Hurler syndrome. At 1 month post-HSCT, peripheral donor myeloid chimerism was >75% and >99% for 0.3x106 and 10x106 cells, respectively. In the brain, transplant of 10x106 cells led to significantly higher donor microglial engraftment versus 0.3x106 cells (Figure A). Notably, high cell dose resulted in significantly higher levels of IDUA enzyme in the brain (Figure B), reduced levels of β-hexosaminidase and glycosaminoglycan (GAG) substrate, and normalization of behavioral outcomes, including rotarod performance, to wild type levels (Figure C). In peripheral tissues, transplant of 10x106 cells, but not 0.3x106 cells, led to a reduction of GAGs to wild type levels as early as 1 week post-HSCT (p<0.01). To determine if donor engraftment in the brain is required for cross-correction, we transplanted 10x106 cells into mice conditioned with a myeloablative dose of treosulfan, which is not sufficient to condition the brain for microglia engraftment. Treosulfan conditioning, followed by high dose HSCT, led to >99% donor myeloid chimerism in the periphery but neither increased microglial levels nor corrected CNS defects (Figures A-C), suggesting that donor engraftment in the brain is required for disease modification. Long-term outcomes and impact on skeletal phenotype in this model will also be presented. Conclusions . We demonstrate that high dose HSCT leads to robust microglia engraftment in the brain and improved disease endpoints. These data suggest that strategies to increase cell dose, such as MGTA-456, may accelerate resolution of neurologic disease in patients with IMDs. Similar approaches, possibly coupled with gene modification technologies, could be used to improve microglial function in other neurodegenerative diseases where defective microglia have been implicated. Disclosures Goncalves: Magenta Therapeutics: Employment, Equity Ownership, Patents & Royalties. Hyzy:Magenta Therapeutics: Employment, Equity Ownership. Brooks:Magenta Therapeutics: Employment, Equity Ownership. Boitano:Magenta Therapeutics: Employment, Equity Ownership, Patents & Royalties. Cooke:Magenta Therapeutics: Employment, Equity Ownership, Patents & Royalties.

Cardiovascular disease (CVD) is the leading cause of death worldwide. Recently, age-related clonal hematopoiesis (CH) has been recognized as a risk factor for CVD of comparable magnitude to smoking, hypertension and hypercholesteremia. While these other risk factors can be mitigated by pharmacological intervention or lifestyle changes, there are no such strategies in place for CH. As CH is initiated by mutations in hematopoietic stem cells (HSCs), a hematopoietic stem cell transplantat (HSCT) could serve as a curative therapy. However, stem cell transplantation is associated with significant toxicity due in part from current conditioning regimens. There is also no evidence that depletion of the disease-driving clones impacts established atherosclerosis. We developed an antibody drug conjugate (ADC) targeting murine CD45. In the context of stem cell transplantation, the CD45-ADC efficiently depletes endogenous HSCs as well as mature leukocytes while enabling rapid engraftment of an infused stem cell graft. In addition, the CD45-ADCs are not based on broad-acting genotoxic agents that lead to long-lasting health risks. We decided to test if CD45-ADC and HSCT could halt atherosclerosis progression through elimination Tet2 knockout HSCs and their disease propagating myeloid progeny. To model CH associated atherosclerosis, LDLR knockout mice were transplanted with 20% CFP labeled wild-type (WT) or Tet2 knockout bone marrow. A single dose of isotype- or CD45-ADC was delivered after 6 weeks of atherosclerosis development and was followed by an infusion of WT CD45.1 bone marrow. As has been reported before, we observed in the isotype-ADC treated animals that Tet2 deficiency leads to a competitive advantage over WT cells. Tet2 knockout cells contributed to peripheral blood chimerism at successively increasing levels and mice harboring the knockout graft showed a significant expansion of their HSC population. Despite their obvious advantage, Tet2 deficient HSC were as efficiently depleted as their WT counterparts upon CD45-ADC and HSCT. Peripheral blood and bone marrow chimerism were similar in WT and Tet2 knockout hosts and the expanded HSC pool was successfully curbed 6 weeks following the intervention. More importantly, CD45-ADC also depleted cells in the atherosclerotic plaques as efficiently as in blood in both WT and Tet2 mutant recipients. This resulted in a significant reduction of myeloid cell infiltration in CD45-ADC conditioned and transplanted knockout hosts and ultimately lead to drastically reduced plaque size in these animals. In conclusion, these data demonstrate that CD45-ADC and HSCT efficiently replaces the disease driving myeloid cells in the atherosclerosis plaques leading to an overall reduction in disease burden. CD45-ADC and transplantation may thus offer a novel therapy for CH and its associated morbidities. Disclosures Palchaudhuri: Magenta Therapeutics: Current Employment. Hyzy:Magenta Therapeutics: Current Employment, Current equity holder in publicly-traded company. Proctor:Magenta Therapeutics: Current Employment. Gillard:Magenta Therapeutics: Current Employment. Boitano:Magenta Therapeutics: Ended employment in the past 24 months, Patents & Royalties. Cooke:Magenta Therapeutics: Ended employment in the past 24 months. Scadden:Magenta Therapeutics: Consultancy, Current equity holder in publicly-traded company, Membership on an entity's Board of Directors or advisory committees.

Resetting the immune system through autologous hematopoietic stem cell transplant (autoHSCT) is a highly effective treatment in patients with autoimmune diseases (AID). AutoHSCT achieved long-term remission in patients with relapsed refractory and secondary progressive multiple sclerosis (Muraro 2017), superior to their previous standard of care (Burt 2019). AutoHSCT in scleroderma patients achieved superior outcomes in two randomized studies (Tyndall 2014, Sullivan 2018). These impressive results are achieved by both eradication of autoreactive immune effector cells and re-establishment of a self-tolerant immune system, i.e., immune system reset. However, only a fraction of eligible patients undergo autoHSCT, in part due to toxicity associated with current conditioning regimens that remove the disease-causing cells.To enable more patients to benefit from immune reset without debilitating chemotherapy, we generated novel anti-human CD45 ADCs that cross react with NHP and evaluated these for the ability to deplete hematopoietic cells in vitro and in vivo (Fig.1). In vitro the CD45-ADC efficiently killed human CD34+ progenitors and peripheral CD3+ T cells from both healthy donors and MS patients. In vivo in humanized NSG mice, single doses of the CD45-ADCs were well-tolerated and led to depletion of human hematopoietic cells in BM (Fig. 1A) and periphery. In NHPs, single doses of CD45-ADCs were well tolerated and depleted BM HSCs and peripheral lymphocytes (Fig.1B). CD45-ADC treatment of hNSGs with sclerodermatous xenoGVHD resulted in resolution of symptoms (Fig. 1C).To model the complete immune reset approach in mouse models of AID, we generated a tool anti-mouse CD45 ADC. A single-dose enabled full myeloablation (>99% depletion of LT-HSCs) and complete donor chimerism with congenic HSCT (>90% chimerism at 16 weeks). In an adoptive transfer model of type I diabetes, treatment with a single dose of CD45-ADC and congenic HSCT led to disease prevention. In a murine model of MS EAE, a single dose of the CD45-ADC followed by congenic HSCT enabled full donor chimerism; treatment prior to disease onset significantly delayed disease onset and reduced disease severity; treatment after disease onset also halted progression of symptoms. These data demonstrate that CD45-ADC conditioning followed by congenic HSCT is sufficient for full myeloablation and immune reset. Experiments are ongoing, and evaluation of this ADC in murine models of diabetes and arthritis will be presented.These results indicate that targeted immune depletion with a single dose of CD45-ADC may be sufficient to enable auto-HSCT and immune reset in multiple AID indications. Targeted conditioning with CD45-ADC may enable more patients to benefit from immune reset through removal of pathogenic cells and autoHSCT without the morbidity and mortality associated with current chemotherapeutic conditioning. Resetting the immune system through autologous hematopoietic stem cell transplant (autoHSCT) is a highly effective treatment in patients with autoimmune diseases (AID). AutoHSCT achieved long-term remission in patients with relapsed refractory and secondary progressive multiple sclerosis (Muraro 2017), superior to their previous standard of care (Burt 2019). AutoHSCT in scleroderma patients achieved superior outcomes in two randomized studies (Tyndall 2014, Sullivan 2018). These impressive results are achieved by both eradication of autoreactive immune effector cells and re-establishment of a self-tolerant immune system, i.e., immune system reset. However, only a fraction of eligible patients undergo autoHSCT, in part due to toxicity associated with current conditioning regimens that remove the disease-causing cells. To enable more patients to benefit from immune reset without debilitating chemotherapy, we generated novel anti-human CD45 ADCs that cross react with NHP and evaluated these for the ability to deplete hematopoietic cells in vitro and in vivo (Fig.1). In vitro the CD45-ADC efficiently killed human CD34+ progenitors and peripheral CD3+ T cells from both healthy donors and MS patients. In vivo in humanized NSG mice, single doses of the CD45-ADCs were well-tolerated and led to depletion of human hematopoietic cells in BM (Fig. 1A) and periphery. In NHPs, single doses of CD45-ADCs were well tolerated and depleted BM HSCs and peripheral lymphocytes (Fig.1B). CD45-ADC treatment of hNSGs with sclerodermatous xenoGVHD resulted in resolution of symptoms (Fig. 1C). To model the complete immune reset approach in mouse models of AID, we generated a tool anti-mouse CD45 ADC. A single-dose enabled full myeloablation (>99% depletion of LT-HSCs) and complete donor chimerism with congenic HSCT (>90% chimerism at 16 weeks). In an adoptive transfer model of type I diabetes, treatment with a single dose of CD45-ADC and congenic HSCT led to disease prevention. In a murine model of MS EAE, a single dose of the CD45-ADC followed by congenic HSCT enabled full donor chimerism; treatment prior to disease onset significantly delayed disease onset and reduced disease severity; treatment after disease onset also halted progression of symptoms. These data demonstrate that CD45-ADC conditioning followed by congenic HSCT is sufficient for full myeloablation and immune reset. Experiments are ongoing, and evaluation of this ADC in murine models of diabetes and arthritis will be presented. These results indicate that targeted immune depletion with a single dose of CD45-ADC may be sufficient to enable auto-HSCT and immune reset in multiple AID indications. Targeted conditioning with CD45-ADC may enable more patients to benefit from immune reset through removal of pathogenic cells and autoHSCT without the morbidity and mortality associated with current chemotherapeutic conditioning. Figure 1.