SENTI-202, a Selective, Off-the-Shelf, Preclinical CAR-NK Cell Therapy with CD33 and/or FLT3 Activating CAR, Healthy Cell Protection from Endomucin (EMCN) Inhibitory CAR and Calibrated Release IL-15 for Hematologic Malignancies Including AML

Brian S. Garrison, PhD1, Han Deng, PhD1, Gozde Yucel, PhD1, Nicholas W. Frankel, PhD1, Russell Gordley, PhD1, Michelle Hung, PhD1, Derrick Lee1, Marcus Gainer1, Assen Roguev1, Wesley Gorman1, Nelia Leemans, MS1, Alice Lam1, Yongshuai Li, MS1, Wilson Wong, PhD2, Kanya Rajangam, MD, PhD1, Philip Lee, PhD1, Timothy Lu, MD, PhD1 1Senti Biosciences, Inc. South San Francisco, CA; 2Department of Biomedical Engineering, Boston University, Boston, MA ASH Annual Meeting 2022, New Orleans, LA Abstract# 1978 SENTI-202, a Selective, Off-the-Shelf, Preclinical CAR-NK Cell Therapy with CD33 and/or FLT3 Activating CAR, Healthy Cell Protection from Endomucin (EMCN) Inhibitory CAR and Calibrated Release IL-15 for Hematologic Malignancies Including AML SENTI-202 Possesses Significant Preclinical Anti-Tumor ActivitySENTI-202 is an Allogeneic Logic-Gated Preclinical CAR-NK Cell Therapy Engineered to Address Unmet Needs in Hematopoietic Malignancies, Including AML SENTI-202 Inhibitory CAR Design and Function SENTI-202: All Genetic Components Delivered as a Single Transduction This project has been funded in whole or in part with Federal funds from the National Cancer Institute, National Institutes of Health, Department of Health and Human Services, under Contract No. 75N91021C00026 oSENTI-202, a preclinical off-the-shelf CAR-NK cell therapy, is engineered with a CD33 OR FLT3 NOT EMCN logic-gated gene circuit and crIL-15 for the treatment of patients with hematologic malignancies, including AML oSENTI-202 has demonstrated: oSignificant in vitro cytotoxicity against leukemia cell lines, andAML LSCs and blast cells from primary patient tumor samples oSignificant in vivo activity within an AML (MV4-11) xenotransplantation model oSignificant in vitro protection of EMCN+ CD33+ FLT3+ model healthy cells, which would otherwisebe killed in the absence of a NOT gate oSignificant increase in vivo of killing precision by demonstrating robust on-tumor killing of EMCN-CD33+ FLT3+ cells with concurrent protection and expansion of EMCN+ CD33+ FLT3+ healthy model cells oSignificant protection of primary healthy human EMCN+ HSCs from off-tumor toxicity Clinical evaluation of SENTI-202 is planned (IND expected in 2023) to evaluate the safety and efficacy in patients with hematologic malignancies with high unmet need, including AML. LIR1 Identified as Optimal iCAR ICD for SENTI-202 Summary Identification of optimal iCARintracellular domain (ICD) for SENTI-202. (A.) We developed a systematic approach to identify the optimal SENTI-202 iCARICD using a screen of native inhibitory proteins in the context of an anti-EMCN iCAR, which were co-expressed and assayed for function in concert with FLT3 aCARs. Constructs were evaluated based on HIGH on-tumor (FLT3+EMCN-) killing and LOW off-tumor (FLT3+EMCN+) killing. Killing was CAR- mediated (basal NK cytotoxicity subtracted). (B.) Comparative analyses between NOT gates with different iCARICDs identified LIR1 as the optimal ICD for SENTI-202. SENTI-202 possesses robust killing activity against leukemia cell lines (SEM, MV4-11), which lack EMCN expression. However, once these cells are engineered to express EMCN, the SENTI- 202 EMCN iCAR then provides significant protection from aCAR-mediated cytotoxicity.p-value: ***=0.0001-0.001, ****=<0.0001 EMCN iCAR Protects EMCN+ Cells In Vivo:SENTI-202 iCAR Protects Cells Expressing the Healthy Cell Marker EMCN The EMCN NOT gate enables SENTI-202 to avoid on-target off-tumor toxicity in vivo. (A.) Thein vivo mixed target study design. “On-Tumor” target cells (EMCN-FLT3+ CD33+ cancer model) and “Off-Tumor” target cells (EMCN+ FLT3+ CD33+ healthy model) were mixed with NK cells and injected into mice. (B.) Peripheral blood flow cytometry analysis from representative mice showing selective killing of On-Tumor cells by SENTI-202. (C.)The SENTI-202 NOT gate enables specific protection and enrichment of Off-Tumor target cells (EMCN+ healthy model), compared to the OR gate-only control. (D.) Abundance of target cells of each type relative to all CD45+ blood cells showing that On-Target Tumor cells are reduced by NK cells (especially in OR gate only and SENTI-202 groups), but notably, “Off-Tumor” healthy target cells are only spared by SENTI-202 since they possess the EMCN iCAR (NOT gate). p-value: *=0.01-0.05, **=0.001-0.01, ****=<0.0001 SENTI-202: Mechanism of Action SENTI-202 demonstrates significantin vitro CAR-dependent anti-tumor cytotoxicity against (A.) SEM leukemia cells which express CD33 and FLT3, and (B.) AML LSCs and AML blasts from primary patient samples. (C.) SENTI-202 cell therapy shows significant in vivo activity within an AML (MV4-11) xenotransplantation model. p-value: *=0.01-0.05, **=0.001-0.01, ***=0.0001-0.001, ****=<0.0001 SENTI-202 Demonstrates In Vitro and In Vivo Anti-Tumor Activity (A.) SENTI-202 CAR-NK cells possess similar cytotoxicity potential as OR gate-only control CAR-NK cells against leukemia target cells. However, (B.) only the SENTI-202 cells (with the EMCN iCAR) were able to provide significant protection to the healthy EMCN+ HSCs. p-value: **=0.001-0.01 Potential Need for NOT GATE for Increased Healthy HSC Protection HSCs Express FLT3 Identification of the Healthy HSC Marker EMCN EMCN iCAR Protects Healthy Human HSCs from aCAR-Mediated Killing PBS Non-Engineered NK Cells SENTI-202Day 7 21 28 34 Treatment Groups AML #1 (M1) Blast LSC ** *** AML #2 (M4) Blast LSC **** *** AML #3 (M2) Blast LSC ** * SENTI-202 includes a NOT GATE (iCAR) to protect healthy HSCs from potential off-tumor toxicity. (A.) The tumor-associated targetantigen FLT3 is expressed on both AML LSCs and healthy HSCs, which necessitates the use of an iCAR to help protect healthy HSCs from potential off-tumor toxicity. (B.) We identified EMCN as a potential HSC-specific iCAR target using a multi-step bioinformatics pipeline, and validated EMCN expression as a healthy HSC surface marker that is not expressed on AML cells using (C.) transcriptomics and (D.) flow cytometry.p-value: ****=<0.0001 A. B. C. D. 0 20 40 60 80 100 Non-engineered NK cells SENTI-202 Percent Cytotoxicity 0 20 40 60 80 100Percent Cytotoxicity 1:4 1:2 1:1 2:1 4:1 Effector:Target (E:T) Ratio Non-engineered NK cells SENTI-202 A. B. C. *** *** **** ******** AML blastsAML LSCs Healthy HSCsNeutrophils Log2microarray expression 4 6 8 10 12 14 FLT3 Expression EMCN Expression Log2microarray expression EMCN Protein A. B. 0 20 40 60 80Percent aCAR-Mediated Cytotoxicity SEM MV4-11MV4-11SEM (1:2; E:T) (1:4; E:T)(1:1; E:T) (1:2; E:T) EMCN-target cell EMCN+target cell **** **** *** *** >60% aCAR + iCAR Co-Expression crIL-15 EMCN inhibitory CAR (iCAR) CD33/FLT3 activating CAR (aCAR) The SENTI-202 gene circuit encodes for 3 cargos: (1.) crIL-15, (2.) bivalent CD33/FLT3 aCAR, and (3.) EMCN iCAR, that are expressed in a tricistronicformat (see above) that enables (A.) up to 63.2% co-expression of the bivalent CD33/FLT aCAR and EMCN iCAR, (B.) IL-15 production of upto 107pg/ml/1e6, and (C.) up-regulation of the NK cell IL-15 receptor signaling cascade via pSTAT5. note: 2A=2A self-cleavage peptide. p-value: *=0.01-0.05, ***=0.0001-0.001 2A 2ASENTI-202 gene circuit IL-15 Expression and NK Cell Activation IL-15 (pg/ml/1e6 cells) MFI of pSTAT5 NK cells NK cells SENTI-202 SENTI-202 Bivalent CD33/FLT3 aCAREMCN iCAR aCAR + iCAR 63.2% A. B. C. Background: Patients with CD33+ and/or FLT3+ malignancies, which encompass most myeloid malignancies such as acute myeloid leukemia (AML), have a grim prognosis and a high unmet need. There are no approved cell therapies for patients with AML due to paucity of targets and an immunosuppressive bone marrow milieu. SENTI-202 is a preclinical CAR-NK cell product engineered with a CD33 and/or FLT3 NOT EMCN logic-gated gene circuit and a proprietary calibrated release IL-15 (crIL-15) technology to overcome these challenges. The bivalent CD33/FLT3 activating CAR (aCAR), EMCN inhibitory CAR (iCAR) and crIL-15 are all engineered within a single gene circuit onto a natural killer (NK) cell (healthy adult peripheral blood-derived)to create an off-the-shelf chimeric antigen receptor (CAR) NK cell therapy. While previous presentations (Garrison et al., ASH, 2021) focused on the performance of individual SENTI-202 components, this presentation focuses on functional data from the complete logic-gated gene circuit and the crIL-15 together in an NK cell (i.e., the final CAR-NK cell with all 3 genetic elements for clinical development). OR GATE: bivalentCD33 and/or FLT3 activation → designed to target both AML LSCs and AML blasts to enable the potential for deep and durable responses in acute myeloid leukemia (AML) and other blood cancers NOT GATE: inhibition by EMCN protective antigen selectively expressed on healthy cells → potential for improved safety and increased therapeutic window In Vivo Assessment of NOT logic gating AML blastsAML LSCs Healthy HSCsNeutrophils OR/NOT Logic Gate Enables Precise On-Tumor Cell Killing NOT Gate Protects Off-Tumor Cells Precise On-Tumor Killing with Off-Tumor Protection SENTI-202 SENTI-202 SENTI -202 SENTI -202 SENTI -202 Off-Tumor Target: Healthy EMCN+ HSCs On-Tumor Target: Leukemia cell line aCAR-mediated killing (%) aCAR-mediated killing (%) OR Gate-only CAR-NK SENTI-202 OR Gate-only CAR-NK SENTI-202 ** ns 0 20 40 60 0 20 40 60 A. B. *** * crIL-15→ potential for increased persistence, autocrine and paracrine immune cell activation SENTI’S LOGIC GATES SOLVE KEY DISEASE CHALLENGES IN AML Target Heterogeneity Limited efficacy due to incomplete targeting of leukemic blasts and leukemic stem cells (LSCs) and on-target/off- tumor toxicity due to AML Tumor-Associated Antigens (TAAs) being expressed on both tumor and healthy cells OR + NOT GATE Targets multiple AML TAAs (OR GATE) and enables targeting of both AML blasts (CD33) and LSCs (FLT3) while preserving healthy blood stem cells (NOT GATE) CHALLENGE SENTI GENE CIRCUIT SOLUTION 29.5% 5-Year Survival at Diagnosis1 UNMET NEED IN AML Diseased bone marrow Insufficient targeting of AML blasts and LSCs lead to relapse Overall survival in R/R AML patients ~5-10 Months1 Selective killing A. B. C. D. 1 SEER Cancer Stat Facts: Acute Myeloid Leukemia Target cells with respective Effector:Target(E:T) cell ratio

The release as fetched from its publisher. sentibio.com ↗