sana-10k_20211231.htm

UNITED STATES

SECURITIES AND EXCHANGE COMMISSION

Washington, D.C. 20549

FORM 10-K

(Mark One)

For the fiscal year ended December 31, 2021

Commission File Number 001-39941

Sana Biotechnology, Inc.

(Exact name of Registrant as specified in its Charter)

188 East Blaine Street, Suite 400 Seattle, Washington 98102

(Address of principal executive offices) (Zip Code)

Registrant’s telephone number, including area code: (206) 701-7914

Securities registered pursuant to Section 12(b) of the Act:

Title of each class Trading Symbol(s) Name of each exchange on which registered

Common Stock, $0.0001 par value SANA Nasdaq Global Select Market

Securities registered pursuant to Section 12(g) of the Act: None

Indicate by check mark if the Registrant is a well-known seasoned issuer, as defined in Rule 405 of the Securities Act. Yes☒ No ☐

Indicate by check mark if the Registrant is not required to file reports pursuant to Section 13 or 15(d) of the Act. Yes ☐No☒

Indicate by check mark whether the Registrant: (1) has filed all reports required to be filed by Section 13 or 15(d) of the Securities Exchange Act of 1934 during the preceding 12 months (or for such shorter period that the Registrant was required to file such reports), and (2) has been subject to such filing requirements for the past 90 days. Yes☒ No ☐

Indicate by check mark whether the Registrant has submitted electronically every Interactive Data File required to be submitted pursuant to Rule 405 of Regulation S-T (§232.405 of this chapter) during the preceding 12 months (or for such shorter period that the Registrant was required to submit such files). Yes☒ No ☐

Indicate by check mark whether the registrant is a large accelerated filer, an accelerated filer, a non-accelerated filer, smaller reporting company, or an emerging growth company. See the definitions of “large accelerated filer,” “accelerated filer,” “smaller reporting company,” and “emerging growth company” in Rule 12b-2 of the Exchange Act.

Large accelerated filer ☐ Accelerated filer ☐

Non-accelerated filer ☒ Smaller reporting company ☐

Emerging growth company ☒

If an emerging growth company, indicate by check mark if the registrant has elected not to use the extended transition period for complying with any new or revised financial accounting standards provided pursuant to Section 13(a) of the Exchange Act. ☐

Indicate by check mark whether the registrant has filed a report on and attestation to its management’s assessment of the effectiveness of its internal control over financial reporting under Section 404(b) of the Sarbanes-Oxley Act (15 U.S.C. 7262(b)) by the registered public accounting firm that prepared or issued its audit report. ☐

Indicate by check mark whether the Registrant is a shell company (as defined in Rule 12b-2 of the Exchange Act). Yes ☐ No ☒

The aggregate market value of the voting and non-voting common equity held by non-affiliates of the Registrant was approximately $3.7 billion, based on the closing price of the Registrant’s common stock on The Nasdaq Global Select Market LLC on June 30, 2021, the last business day of the Registrant’s most recently completed second fiscal quarter. Shares of the Registrant’s common stock held by each officer and director and stockholders that the Registrant has concluded are affiliates of the registrant have been excluded in that such persons may be deemed affiliates of the Registrant. This determination of affiliate status is not a determination for other purposes.

The number of shares of Registrant’s common stock outstanding as of March 11, 2022 was 189,543,388.

DOCUMENTS INCORPORATED BY REFERENCE

Portions of the Registrant’s definitive Proxy Statement relating to its 2022 Annual Meeting of Stockholders (Proxy Statement) are incorporated by reference into Part III of this Annual Report on Form 10-K (Annual Report) where indicated. The Proxy Statement will be filed with the U.S. Securities and Exchange Commission within 120 days after the end of the fiscal year to which this Annual Report relates.

Table of Contents

Page

PART I

Item 1. Business 5

Item 1A. Risk Factors 80

Item 1B. Unresolved Staff Comments 138

Item 2. Properties 138

Item 3. Legal Proceedings 138

Item 4. Mine Safety Disclosures 138

PART II

Item 6. [Reserved] 140

Item 7A. Quantitative and Qualitative Disclosures About Market Risk 154

Item 8. Financial Statements and Supplementary Data 155

Item 9A. Controls and Procedures 181

Item 9B. Other Information 181

Item 9C. Disclosure Regarding Foreign Jurisdictions that Prevent Inspections 181

PART III

Item 10. Directors, Executive Officers and Corporate Governance 182

Item 11. Executive Compensation 182

Item 14. Principal Accounting Fees and Services 182

PART IV

Item 15. Exhibits, Financial Statement Schedules 183

SPECIAL NOTE REGARDING FORWARD-LOOKING STATEMENTS

This Annual Report on Form 10-K (Annual Report) contains forward-looking statements that involve substantial risks and uncertainties. All statements other than statements of historical facts contained in this Annual Report could be deemed forward-looking statements, including those statements highlighted below. In some cases, you can identify these statements by forward-looking words such as “aim,” “anticipate,” “believe,” “continue,” “could,” “estimate,” “expect,” “intend,” “may,” “might,” “plan,” “potential,” “predict,” “should,” “would,” or “will,” the negative of these terms, and other comparable terminology. These forward-looking statements, which are subject to risks, include, but are not limited to, statements about:

• our clinical and regulatory development plans;

• the pricing and reimbursement of our product candidates, if approved;

• the performance of our third-party suppliers and manufacturers;

• our future financial performance;

We have based these forward-looking statements largely on our current expectations, estimates, forecasts, and projections about future events and financial trends that we believe may affect our financial condition, results of operations, business strategy, and financial needs. In light of the significant uncertainties in these forward-looking statements, you should not rely upon forward-looking statements as predictions of future events. Although we believe that we have a reasonable basis for each forward-looking statement contained in this Annual Report, we cannot guarantee that the future results, levels of activity, performance, or events and circumstances reflected in the forward-looking statements will be achieved or occur at all. You should refer to the sections titled “Risk Factors” and “Management’s Discussion and Analysis of Financial Condition and Results of Operations” for a discussion

of important factors that may cause our actual results to differ materially from those expressed or implied by our forward-looking statements. Other sections of this Annual Reportmay include additional factors that could harm our business and financial performance. New risk factors emerge from time to time, and it is not possible for our management to predict all risk factors, nor can we assess the impact of all factors on our business or the extent to which any factor, or combination of factors, may cause actual results to differ materially from those contained in, or implied by, any forward-looking statements. Except as required by law, we undertake no obligation to publicly update any forward-looking statements, whether as a result of new information, future events, or otherwise.

RISK FACTOR SUMMARY

Investing in our securities involves a high degree of risk. Below is a summary of material factors that make an investment in our securities speculative or risky. Importantly, this summary does not address all of the risks that we face. Additional discussion of the risks summarized in this Risk Factor Summary, as well as other risks that we face, can be found under the heading “Risk Factors” in Part I of this Annual Report.

Our business is subject to a number of risks of which you should be aware before making a decision to invest in our common stock. These risks include, among others, the following:

PART I

Item 1. Business.

Overview

We were founded on the belief that engineered cells will be one of the most important transformations in medicine over the next several decades. The burden of diseases that can be addressed at their root cause through engineered cells is significant. We view engineered cells as having the potential to be as therapeutically disruptive as biologics to clinical practice. Our long-term aspirations are to be able to control or modify any gene in the body, to replace any cell that is damaged or missing, and to markedly improve access to cellular and gene-based medicines. We have brought together an experienced group of scientists, engineers, and company builders and combined them with the necessary technologies to move this vision forward. We are developing ex vivo and in vivo cell engineering platforms to revolutionize treatment across a broad array of therapeutic areas with unmet treatment needs, including oncology, diabetes, central nervous system (CNS) disorders, cardiovascular diseases, and genetic disorders, among others. Our platform progress, broad capabilities, and strong balance sheet enable us to execute on a broad vision, with a goal of submitting our first INDs in 2022, with the opportunity to submit multiple INDs per year beyond 2022.

We believe the time is right to develop engineered cell therapies across a broad range of therapeutic areas. The field has seen initial clinical proof of concept for gene and cell replacement approaches across multiple diseases, including cancer and certain genetic disorders, through the application of adeno-associated virus (AAV) based gene therapies, autologous CAR T cell therapies, and autologous and allogeneic grafts/transplants. While such existing approaches have limitations, they provide evidence that a broad range of ex vivo and in vivo engineered cells can have transformative clinical potential in at least a subset of patients. Substantial progress in the understanding of genetics, gene editing, gene control, protein engineering, stem cell biology, immunology, process analytics, and computational biology have converged to create an opportunity to markedly increase the breadth and depth of the potential impact of genetic and cellular medicines.

We are seeking to overcome these existing limitations of gene and cell therapy through our ex vivo and in vivo cell engineering platforms, both of which may facilitate the development of therapies that can transform the lives of patients by repairing cells in the body when possible and replacing them when needed. For ex vivo therapies, where diseased cells are damaged or missing entirely and an effective therapy needs to replace the entire cell, a successful therapeutic requires large-scale manufacturing of cells that engraft, function, and persist in the body. Of these, we view persistence as the greatest limitation to dramatically expanding the impact of this class of therapeutics. We believe that product candidates developed with our ex vivo cell engineering platform, which utilizes hypoimmune allogeneic cells that can “hide” from the patient’s immune system, can address this fundamental limitation and unlock a wave of disruptive therapeutics. For in vivo therapies, where the desire is to repair and control genes in the body, a successful product candidate requires both gene modification and in vivo delivery of the therapeutic payload. Of these, we view effective in vivo delivery as the greatest limitation to dramatically expanding the impact of this class of therapeutics. To this end, our initial focus is on cell-specific delivery as well as increasing the diversity and size of payloads.

We believe we have the potential to develop transformative engineered cells as medicines because of our people and our capabilities:

Our people are the most important strength of the company. We have assembled a diverse group of experienced company builders, scientists, manufacturing scientists, engineers, and operators to execute our business plan.

Our capabilities enable us to take a comprehensive approach to the most important and difficult aspects of engineering cells. We are pursuing ex vivo and in vivo cell engineering and can leverage the synergistic proficiencies required to succeed in both approaches. We believe we can capitalize on the shared expertise and infrastructure between the platforms to maximize the potential success and the reach of our transformative therapies. We have built deep internal capabilities across a wide range of areas focused on solving the most critical limitations in engineering cells including:

Our ex vivo and in vivoCell Engineering Platforms

The advent of recombinant DNA technology in the 1970s ushered in a new era of therapeutics, enabling the synthetic manufacture of human protein therapies at scale for the first time. However, the critical inflection point occurred when key technological advancements eventually enabled the broad development of monoclonal antibodies with suitable therapeutic properties. These advancements, combined with progress in understanding disease biology, allowed biologics to become the second largest therapeutic class. We believe engineered cells are at a similar inflection point, with key recent technological advancements providing the potential for the broad applicability of this therapeutic class.

Ex vivo cell engineering

Engineering cells ex vivo requires the ability to engineer and manufacture cells at scale and then deliver them to the patient, so that they engraft, function appropriately, and have the necessary persistence in the body. Our goal for ex vivo cell engineering is to replace any cell in the body with cells that engraft, function, and persist over time, and to manufacture those cells cost-effectively at scale. Our ex vivo cell engineering platform utilizes our hypoimmune technology to create cells that can “hide” from the patient’s immune system to enable persistence of allogeneic cells. We are striving to make therapies utilizing pluripotent stem cells with our hypoimmune genetic modifications as the starting material, which we then differentiate into a specific cell type, such as a pancreatic beta cell, before treating the patient. Additionally, for cell types for which effective differentiation protocols from a stem cell have not yet been developed, such as T cells, instead of starting from a pluripotent stem cell, we can utilize a fully differentiated allogeneic cell,

sourced from a donor, as the starting material to which we then apply our hypoimmune genetic modifications. Our goal is to manufacture genetically modified cells that are capable of both replacing the missing cell and evading the patient’s immune system. We are now applying our technologies to make cell products for the treatment of multiple diseases.

In vivo cell engineering

Engineering cells in vivo requires the development of both an appropriate delivery vector as well as a payload to effectively modify the cell. Our goal for in vivo cell engineering is to repair and control the genes of any cell in the body. The ultimate aim is to achieve the delivery of any payload, to any cell, in a specific and repeatable way. Our in vivo cell engineering platform harnesses fusogen technology, which targets cell surface receptors, and thereby can enable cell specific delivery for a meaningful number of different cell types. Using our fusogen technology, we have shown in preclinical studies that we can specifically target numerous cell surface receptors that, when combined with delivery vehicles to form fusosomes, allow cell-specific delivery across multiple different cell types.

Our Portfolio Strategy

We believe the potential applications of our platforms are vast. To prioritize programs for our ex vivo and in vivo engineering pipeline we have used the following strategies:

• prioritize efforts where success in one area begets success in others.

Our Pipeline

We are developing a broad pipeline of product candidates focused on creating transformative ex vivo and in vivo engineered cell therapies across a range of therapeutic areas. We are in the early stages of development across a broad pipeline of product candidates, all of which are currently in the preclinical stage of development and are summarized below:

Each of our initial programs provides the potential for meaningful standalone value while also supporting our potential ability to further exploit our platforms into broadly applicable medicines.

Our most advanced hypoimmune product candidate is SC291, a CD19-directed allogeneic CAR T program for NHL, CLL, and ALL. This program is designed to address the major limitation of existing allogenic CAR T cell therapies: evasion of host versus

graft responses (HvGR), which occurs when a patient’s immune system kills the transplanted T cells, limiting the potential benefit of the therapy. One approach to avoid HvGR has been to effectively eliminate a patient’s immune system for a short period using chemotherapy, which puts the patient at risk for severe infections. Further, the patient’s immune system will inevitably recover, which will lead to the immune system eliminating the CAR T cells, limiting the effectiveness of the therapy. Our hypoimmune technology is designed to hide cells from the patient’s immune system, giving our allogeneic CAR T cell program the potential to create medicines that persist longer in patients and avoid the risksassociated with higher doses of chemotherapy. Our goal is to submitan IND for SC291 in 2022, and initial clinical success would unlock meaningful standalone value in the development of SC291 in NHL/CLL/ALL which has the potential to address scalability challenges of autologous therapies. Additionally, this initial clinical success would support and validate the expansion of our allogenic CART efforts, which include a CD22-targeting allogeneic CAR T which could be combined with targeting CD19(SC276) and offers the potential benefit of higher and more durable complete response rates, and a BCMA directed CART (SC255) in multiple myeloma. Furthermore, initial clinical success would also support the validation of the hypoimmune platform overall, which is being actively deployed internally across a number of other therapeutic areas beyond oncology.

Our next most advanced hypoimmune product candidate is SC451, PSC-derived pancreatic beta cells for the treatment of diabetes, with an initial focus on Type I diabetes mellitus (T1DM). Almost 1.6 million people in the United States, and 2.4 million in Europe have T1DM. T1DM is a disease in which a patient’s immune system attacks and kills pancreatic beta cells, leading to complete loss of insulin production in affected individuals. Patients need to take multiple insulin injections every day for life, and, while insulin has a profoundly positive impact on patients, people with T1DM have approximately 15 years shorter life expectancies than people without diabetes and are consistently at risk for complications such as coma, stroke, myocardial infarction, kidney failure, and blindness from poorly controlled blood glucose. We and our collaborators have shown that we can develop high quality beta cells that, when transplanted, normalize blood glucose and cure diabetes in animal models. We have also shown that our hypoimmune cells induce no systemic immune response, even in NHPs with a pre-existing immune response to non-hypoimmune cells. As a result, we believe our stem cell derived hypoimmune pancreatic cells have the potential to create a disruptive treatment for T1DM, offering patients life-long normal blood glucose without immunosuppression. We are working through the process development and IND-enabling studies to allow for an IND submission for SC451 as early as 2023.

Our most advanced fusosome product candidate is SG295, which targets CD19+ cancer cells, including NHL, CLL, and ALL. This program provides us with an opportunity to develop potential product candidates to expand access to CAR T cell therapy to many more patients in need. In addition, we believe the ability to deliver a payload encoding a CAR to a T cell without meaningful ex vivo manipulation has the potential to improve effectiveness over ex vivo manufactured CAR T cell products. These approaches should result in the generation of therapeutically active CAR T cells without the complexities and delays associated with the process of T cell collection and ex vivo manufacturing. Furthermore, the ex vivo expansion in the presence of high cytokine concentrations, while necessary for the manufacture of approved CAR T cell products, also contributes to marked changes in T cell quality that may not be therapeutically beneficial. The generation of a CAR T cell within the natural physiological environment in vivo has the potential to improve the quality of the CAR T cell generated, potentially improving both efficacy and the side effect profile. Finally, the effectiveness of ex vivo manufactured CAR T cells is dependent on the administration of a lymphodepleting preparative regimen prior to infusion to facilitate expansion of the CAR T cell product, which can have meaningful adverse safety implications. We do not expect to need a lymphodepleting regimen prior to in vivo delivery of the CAR gene, as our goal is to expose our fusosomes to as many T cells in the body as possible. Our goal is to submit an IND for SG295 in 2022 and initial clinical success would unlock meaningful standalone value in the development of SG295 in NHL/CLL/ALL Additionally, this initial clinical success would support and validate the expansion of our in vivo CAR T efforts and support the validation of our fusosome platform overall, which is being actively deployed internally across a number of other therapeutic areas beyond oncology with the goal of targeted delivery of DNA and gene editing machinery to specific cells in vivo.

Our ex vivo Cell Engineering Platform

Overview

Ex vivo cell engineering aims to treat human disease by engrafting new cells to replace diseased cells that are damaged or missing in patients. Historically there have been four key challenges to ex vivo cell engineering:

• engraftment of the right cell in the right environment;

• manufacturing the desired cell in the quantities required.

Our ex vivo cell engineering platform seeks to address these four challenges and is focused on engineering hypoimmune cells that engraft, function, and persist in patients by evading immune rejection. These are derived from cell sources that are scalable and we believe that continued progress with this platform has the potential to create broad access for patients.

Our Approach to Building our ex vivo Cell Engineering Platform

We have approached the development of our ex vivo cellengineering platform by investing in solutions to address the key challenges outlined above:

Our Approach to Building our ex vivo Cell Engineering Portfolio

We have prioritized cell types for our programs where:

• high unmet need can be addressed by cell replacement;

Based on this prioritization, we are initially focused on four cell types: T Cells, Beta cells, GPCs,and Cardiomyocytes.

Historical context of ex vivo therapy

Blood transfusions have been a standard treatment for many patients for over 100 years. The first successful kidney transplant occurred in 1954, followed by the first successful heart transplant in 1967, demonstrating the transformative clinical potential of replacing damaged or missing cells in the body. Surgical enhancements have improved the success of engraftment, but lack of organ access, a complex surgical procedure, and immune rejection of the donated organs have limited the impact of these procedures.

Progress in immunosuppressive regimens, such as the development of cyclosporine, has improved organ survival rates. However, substantial side effects and the fact that many patients are ineligible or non-compliant have reduced their impact.

Ultimately, the field has looked for a scalable source of therapeutic cells that can be accessed broadly at a manageable cost, as well as cells that can evade immune rejection without immunosuppression. The advent of stem cell technology and subsequent improvements in methods to generate functional differentiated cells at scale have the potential to address the shortage of donor tissues and organs. In addition, over the past decade a deeper understanding of the immunology of host versus graft responses, coupled with novel techniques to manipulate the immunological profile of cells via gene editing, have raised the prospect that ex vivo engineered cells can significantly benefit patients without the requirement for significant immunosuppression.

Sources of allogeneic cells

There are three main potential sources of allogeneic cells, or cells that do not originate from the patient, and therefore have the potential to be manufactured and supplied at scale. These are embryonic stem cells (ESCs), iPSCs, and donor-derived cells. Our portfolio currently reflects a mix of sources, with the ambition of transitioning primarily to iPSCs over time.

Embryonic Stem Cells

The recognition that every cell in the body originates from a zygote, or fertilized egg, led to the research and ultimate discovery of human ESCs, with the derivation of the first human ESC line in 1998. ESCs are pluripotent stem cells which can potentially differentiate into any cell type and are derived from the inner cell mass of a blastocyst or pre-implantation stage embryo. They are typically cultured in vitro and grown through cycles of cell division, known as passages, until a line of cells is established that can proliferate without differentiating, and retain their pluripotency while remaining well characterized, including free from potentially deleterious genetic mutations. Because pluripotent stem cells can divide indefinitely without exhaustion, an ESC line can be used to generate cell banks, consisting of large numbers of well-characterized vials of cells, that can be frozen and stored for future use.

Induced Pluripotent Stem Cells

The discovery that mature, differentiated cells can be reprogrammed to be the equivalent of an ESC and capable of generating any cell type in the body, has led to the research and ultimate development of human iPSCs, providing an alternative option as a source of stem cells for use in ex vivo engineered cells. A key scientific step was the breakthrough in 2006 demonstrating that mature cells could be reprogrammed via the expression of a small number of genes to result in pluripotent cells These iPSCs have similar potential to ESCs to be used as an indefinitely renewable cell bank for manufacturing of cell-based therapies.

Donor-Derived Allogeneic Cells

Another source of cells, which we utilize in our T cell program, comes from mature donor-derived allogeneic cells. While these cells are neither pluripotent nor from an infinitely renewable source, T cells can be obtained as mature cells from human donors at scale. The use of donor-derived cells for our T cell program should allow us to rapidly advance the program towards the clinic with the implementation of our hypoimmune technology.

Approach to Sources of Allogeneic Cells

We are focused on iPSCs as the starting material for our programs, which offers regulatory and cultural advantages to ESCs, and scale and product consistency advantages to donor-derived allogeneic cells. Our portfolio currently reflects a mix of sources, which is primarily driven by historical factors as well as current better characterization of genomic stability through differentiation. Our ambition is to transition primarily to iPSCs over time.

Crucial aspects of developing allogeneic cells from any source include the thorough characterization of the cells, a comprehensive understanding of the global regulatory environment, and an ability to maintain cells under the required conditions, such as current GMP (cGMP), at various stages of the manufacturing processes. We believe our early investment in building capabilities in the science and manufacturing of these cells will increase our likelihood of success. This investment is anticipated to

yield sources of cells suitable for the global clinical development and commercialization of ex vivo engineered cells for a broad patient population, in line with our vision to democratize access.

Background on Immunological Barriers to ex vivo Therapies and Current Limitations

Starting with studies in renal transplantation in the early 1900s, it became clear that there were immunological factors preventing successful transplantation. Initially, it was suspected to be mediated by an antibody response, but in the 1950s it was discovered that cell-mediated immune pathways also play a critical role in transplant rejection.

Further studies established T cells as playing a key role in the host immune response to transplant. T cells belong to the “adaptive” immune system, recognizing and eliminating “non-self” cells via recognition of differences in cell-surface proteins encoded by the major histocompatibility (MHC) locus. There are two types of MHC molecules: MHC class I, expressed on the surface of almost all nucleated cells, and MHC class II, expressed constitutively on professional antigen presenting cells (APC), including macrophages and dendritic cells. Expression of MHC class II is also induced in many additional cells in the context of inflammation. MHC class I molecules typically display peptides on the cell surface from degraded intracellular proteins. Cells display peptides from normal “self” proteins on MHC class I, which typically will not activate an immune response due to a process called tolerance, where the body recognizes these peptides as “self”. However, if a cell displays a peptide from a foreign or mutated protein on MHC class I, for example as a result of a protein mutation, it may result in the activation of a cytotoxic T cell response specific to the peptide-MHC complex via the T cell receptor (TCR) on the T cell surface. The activated T cell then eliminates the cell. MHC class II molecules typically display peptides derived from phagocytosis of extracellular proteins on the surface of APCs. These peptide-MHC complexes interact with TCRs on helper T cells, such as CD4+ T cells, resulting in a downstream cellular and humoral immune response. The humoral immune response leads to antibody production against foreign proteins. In allogeneic transplants, the cellular and humoral processes can recognize proteins from the donor as “foreign”, resulting in an immune response to the transplant including potential elimination of the transplanted cells. In the allogeneic setting, MHC proteins can be highly immunogenic due to their inherent polymorphism, increasing the risk of the recognition of transplants as “foreign”. This underlies the basis for MHC typing and matching to assess and reduce the risk of organ transplant rejection.

Many groups have attempted to engineer cells that can evade the adaptive immune system, typically by downregulating or eliminating expression of MHC molecules on the surface of cells. While this can reduce the adaptive immune response to donor cells, the human immune system has evolved so that parts of the innate immune system will recognize cells missing MHC molecules and eliminate them. For example, natural killer (NK) cells express receptors known as inhibitory killer-cell immunoglobulin-like receptors (inhibitory KIRs). KIRs recognize self MHC class I molecules on the surface of cells and provide inhibitory signals to the NK cells to prevent their activation. Cells missing MHC class I molecules are correspondingly eliminated by NK cells because of the lack of inhibitory KIR signaling and a resulting cytolytic activation. Known as the “missing self-hypothesis,” this important redundancy in immunology enables the elimination of virally infected or transformed cells that have downregulated MHC class I, but also has complicated the development of allogeneic cells as broadly applicable therapeutics. Our hypoimmune technology seeks to engineer cells to avoid immune rejection by addressing both the adaptive and innate immune response.

There are three key strategies that have been utilized to date to overcome immune rejection, with limited success:

Our Solution – Hypoimmune Technology

To address the challenge of immune rejection with allogeneic cell transplantation, we are developing our hypoimmune technology, utilizing gene modification to introduce permanent changes to the cells. We are applying the hypoimmune technology to both iPSCs, which can then be differentiated into multiple cell types, and to donor-derived allogeneic T cells, which has the goal of making potent CAR T cells at scale. Our goal with this technology is to transplant allogeneic cells into patients without the need for systemic immune suppression. We believe that enabling this capability has the potential to enable ex vivo engineered cells to become an important therapeutic modality alongside small molecules, protein biologics, and in vivo engineered cells.

Some of our scientific founders, including Dr. Sonja Schrepfer, our Senior Vice President, Head of Hypoimmune Platform, and their collaborators have worked on creating hypoimmune cells for well over a decade. A key insight was focusing on the phenomenon of fetomaternal tolerance during pregnancy. The fetus, despite having half its genetic material from the father, is not rejected by the mother’s immune system. However, after birth, few if any children would qualify as a matched donor for a cell or organ transplant for their mother. These scientists categorized the differences of the maternal-fetal border and systematically tested them to understand which, if any, of these were most important to immune evasion. They have tested these changes in both in vitro and in vivo animal models.

Designing Hypoimmune Cells

Our goal is to create a universal cell that is able to evade immune detection, regardless of cell type or transplant location. Our first-generation technology, which is progressing through late-stage animal confirmatory studies, combines the three gene modifications below to hide these cells from the host immune system:

• disruption of MHC class I expression;

• disruption of MHC class II expression; and

Preclinical Development of Hypoimmune Cells

We and our licensors have carried out a series of experiments in various model systems of increasing immunological complexity. These included (i) transplanting undifferentiated mouse hypoimmune iPSCs- into MHC mismatched allogeneic mice, (ii) transplanting mouse hypoimmune iPSC-derived differentiated cells, such as endothelial cells, into MHC mismatched allogeneic mice, (iii) transplanting human hypoimmune iPSCs into MHC mismatched humanized allogeneic mice; and (iv) transplanting human hypoimmune iPSCs into non-human primates (NHPs). We are currently carrying out experiments transplanting NHP hypoimmune iPSC cells into NHPs as well as transplanting NHP hypoimmune iPSC-derived differentiated cells, such as cardiomyocytes, into allogeneic NHPs.

Each mouse experiment evaluated:

We are investigating both human iPSCs in NHPs as well as NHP iPSCs in NHPs, as we want insights into how the NHP immune system reacts to each of these species. We have largely completed the study of human iPSCs and have early results from the NHP hypoimmune iPSC transplantation experiments. We are encouraged by data to date across species, with the NHP immune system most closely resembling the human immune system, representing the strictest test outside of testing these cells in humans. We are evaluating both iPSCs as well as differentiated cells transplanted into the microenvironment we intend to target in humans. Based on the results of these NHP studies, we expect to test these hypoimmune cells in humans as a next step.

Mouse iPSC-derived hypoimmune cells transplanted into MHC mismatched allogeneic mouse

Mouse hypoimmune iPSCs transplanted into an MHC mismatched allogeneic mouse were protected from the mouse immune system, and no evidence was seen of either adaptive or innate immune system activation. The control arm transplanted non-edited mouse iPSCs into MHC mismatched allogeneic mice, and, as expected, these non-edited mouse iPSCs were rapidly rejected by the recipient’s immune system with a robust adaptive immune response. In another experiment, the genes that code for MHC class I and MHC class II expression were knocked out. These modifications protected the cells from the recipient mouse’s adaptive immune

system, but NK cells rapidly killed the transplanted cells. These data highlight the importance of making all three gene modifications in order to protect cells from the immune system with an allogeneic transplant.

Next, to ensure that hypoimmune gene modifications protected differentiated cells and that these modifications did not impact the ability of iPSCs to differentiate into various cell types, commonly referred to as pluripotency, it was tested whether the hypoimmune iPSCs cells could be differentiate into three different cell types, function in vivo, and evade the host immune system. The three cell types were cardiomyocytes, endothelial cells, and smooth muscle cells. It was observed that hypoimmune iPSCs could successfully differentiate into all three cell types, the cells functioned in the mouse, and the transplanted cells survived for the full standard observation period with no evidence of immune system activations despite any immune suppression. Differentiated cells derived from non-edited iPSC cells led to immune activation in the host mouse, and they did not survive. These data provide initial proof of concept that iPSCs can be genetically modified, and differentiated into target cells that can engraft, function, and evade the recipient’s immune system following transportation.

Human iPSC-derived hypoimmune cells transplanted into MHC mismatched allogeneic humanized mouse

Having demonstrated the ability of mouse iPSC-derived hypoimmune cells to satisfy each of three testing criteria, the experiments were advanced to evaluate human hypoimmune cells. This was evaluated using a “humanized” mouse system, generated by grafting a functioning human immune system in place of the mouse immune system.

In addition to evaluating the three primary criteria, the ability to successfully engineer human hypoimmune cells from human iPSCs and whether differentiated cells derived from human hypoimmune cells retain biological function were also evaluated.

Creating Hypoimmune Therapeutic Cells from Human iPSCs

Our hypoimmune technology combines the following three gene modifications to hide cells from the host immune system: Disruption of MHC class I and class II expression (which inactivates adaptive immune responses), and overexpression of CD47 (which hides cells from the innate immune system, including macrophages and natural killer (NK) cells). Pluripotent stem cells from healthy donors are used as the starting material and are then genetically modified with the hypoimmune edits. These edited cells are then differentiated into cell types of therapeutic interest, which are administered to the patient as “off the shelf” therapies.

First, the foregoing three edits were replicated in human iPSCs to engineer a human hypoimmune cell line that had comparable properties to the mouse hypoimmune cells in vitro. Next, non-edited human iPSCs were transplanted into MHC mismatched humanized mice. It was observed that these non-edited human iPSCs were rapidly rejected. Human hypoimmune cells were then transplanted into MHC mismatched humanized mice. It was observed that the human hypoimmune cells survived the full length of the experiment and failed to elicit any type of immune response. From this, it was concluded that, in humanized mice, the human

hypoimmune cells can evade the immune system. Pluripotency of human hypoimmune cells was confirmed by differentiation into two different cell types, endothelial cells and cardiomyocytes. These differentiated cells exhibited the characteristics of normal endothelial cells and cardiomyocytes. Finally, to test whether these the differentiated cell types derived from human hypoimmune cells continue to evade the immune system, the differentiated cells were transplanted into humanized mice, and the transplanted cells survived for the full standard observation period. In contrast, differentiated cells derived from non-edited human iPSC cells did not survive after being transplanted, as anticipated. It was also observed that the hypoimmune endothelial cells formed primitive vasculature with active blood flow and the hypoimmune cardiomyocyte cells matured into functional-looking heart cells.

Absence of T and B Cell Activation Following Transplantation of Hypoimmune-Edited Human iPSCs into Mismatched Humanized Mice

Left panels: T cell activation was measured by EliSpot counts for interferon-gamma production. Immune cells from mice that received wild type (wt) iPSC grafts show a brisk interferon response when tested against allogeneic wt iPSC grafts. In contrast, immune cells from mice that received hypoimmune-edited (MHC class I/II disruption, CD47 tg) cells show only minimal interferon production when exposed to allogeneic hypoimmune cells, comparable to background frequency in non-immunized mice. Right panels: B cell activation was measured by antibody binding to each cell type, shown as mean fluorescence intensity (MFI). Wild type cells exhibit significant antibody binding when incubated with serum from mice that received wt cells. In contrast, hypoimmune-edited cells show only background levels of binding when treated with serum from mice that received hypoimmune-edited cells. Adapted from Deuse et al, Nature Biotechnology 2019.

CD47 is Required to Protect Hypoimmune-Edited Cells from Killing by Human NK Cells

Human iPSCs were differentiated into endothelial cells (hiECs) and plated as a monolayer in a multielectrode system. After exposure to NK cells, monolayer viability was measured electrical impedance, indicated here as normalized cell index. As expected, wt cells were not killed by NK cells. In contrast, cells lacking MHC class I and II (but not expressing CD47 tg; MHC class I/II disruption) were rapidly killed. Addition of CD47 tg prevented killing by NK cells. A blocking antibody to CD47 abolished protection from NK cells, affirming the importance of CD47 overexpression in protection from innate immune cell killing. From Deuse et al, Nature Biotechnology 2019.

Survival of Hypoimmune-Edited Human iPSC Grafts in MHC-Mismatched Humanized Mice

Wild type (wt) and hypoimmune-edited (MHC class I/II disruption CD47 tg) iPSCs were engineered to express firefly luciferase before transplantation. Emission of light was used as an index of graft cell viability. Sequential light emission scans from the same representative animal receiving wt cells show progressive loss of graft viability, indicating graft rejection, confirmed quantitatively in the line tracings below. In contrast, mice receiving hypoimmune-edited cells show graft expansion over the course of the experiment, indicating immune evasion. From Deuse et al, Nature Biotechnology 2019.

NHP hypoimmune cells transplanted into NHPs

To evaluate immune evasion properties of the hypoimmune cells, we have tested the immune response to and survival of hypoimmune iPSCs from NHPs by transplantation into an allogeneic NHP recipient without immunosuppression.

Design for Allogeneic Study Involving Wild Type (Unmodified) and Hypoimmune NHP iPSC Delivery to NHPs

The study involved a randomized group of eight NHPs distributed into two cohorts of four NHPs each. The first cohort received an initial intramuscular injection of non-edited NHP iPSCs and a second injection of NHP hypoimmune cells at six weeks (i.e., a crossover design). The second cohort received an initial injection of NHP hypoimmune cells, which allowed assessment of immune

evasion in a naïve recipient. This cohort also received a second injection of non-edited NHP iPSCs, which, with a view towards modeling certain aspects of autoimmune disease, enabled assessment of the impact of injecting hypoimmune cells into an NHP with a pre-existing immune response to non-edited cells. No immunosuppression was administered to any of the animals in the study.

Allogeneic Hypoimmune iPSC Survive in vivo in NHPs with an Intact Immune System

Upper panel: Unmodified wild type (wt) NHP iPSCs (Group 1 top row) or hypoimmune NHP iPSCs (Group 2 bottom row) were introduced via intramuscular injection into allogeneic NHPs. Unmodified NHP iPSCs are undetectable in recipient NHPs by week 3 while hypoimmune NHP iPSCs introduced into naïve NHPs were viable and detectable for 16 weeks post injection. After 6 weeks of the initial injection, NHPs were injected with the cross-over cell type (group 1 with hypoimmune NHP iPSCs and group 2 with wild type iPSCs). In these crossover experiments, hypoimmune NHP iPSCs survived even when the NHP had been exposed to unmodified iPSCs. Unmodified cells injected into NHPs previously injected with hypoimmune iPSCs were rapidly killed with no observable impact on the hypoimmune NHP iPSCs that continued to remain viable. Data are representative for four NHPs receiving HIP iPSCs and wt iPSCs.

Lower panel: iPSC survival is followed over time in vivo using bioluminescence imaging (BLI).

Absence of T Cell, B Cell, or NK Cell Responses Following the First Delivery and Crossover of Hypoimmune NHP iPSCs into NHPs

Upper panel: Immune cells from animals receiving hypoimmune iPSCs showed no response when exposed to hypoimmune iPSCs in vitro (Row 1) in contrast to wt iPSCs (Row 2). Lower panel: Neither unmodified nor hypoimmune-edited cells were susceptible to killing by natural killer (NK) cells, indicating protection from the “missing self” signal. Data above are collected from four NHPs in each experimental arm.

NHP hypoimmune cells grafted into NHPs elicited no detectable systemic immune responses, including no T cell activation and no antibody formation. Innate immune responses mediated by macrophages and NK cells were also undetectable. The transplanted hypoimmune cells were alive and detectable for the duration of the study in these allogeneic recipients (Study duration was 16 weeks for 2/4 NHPs and 8 weeks for 2/4 NHPs. To our knowledge, this is the first instance of prolonged graft survival in an allogeneic transplant setting without immunosuppression in NHPs. In contrast, systemic immune responses from T cells as well as IgM and IgG antibodies were generated to iPSCs without the hypoimmune edits, and the cells were rapidly rejected within two to three weeks.

In the crossover portion of this experiment, injection of NHP hypoimmune cells into NHPs that had previously received non-edited NHPs again elicited no systemic responses as tested in assays for T cell or antibody responses. Similarly, macrophage and NK responses could not be detected. Correspondingly, these cells survived for the full eight weeks that they were monitored suggesting that pre-existing immunity to non-edited human iPSCs had no impact on hypoimmune cell survival. By contrast, in the NHPs that had previously been injected with the hypoimmune cells, the non-edited NHP cells elicited both T cell and antibody responses against the non-edited cells. Notably, these non-edited cells were rapidly rejected (in one to two weeks) in the recipient even as the previously injected hypoimmune cells continued to be viable in the other leg of the NHP. These results provide confirmation that the survival of the hypoimmune allo-graft was not an artifact of an impaired immune system or immune response in the recipient NHP. They also suggest that these hypoimmune cells have the potential for immune evasion even the context of a new immune response toward cells without these edits.

In light of our preclinical data to date, we believe our hypoimmune technology has the potential to address the most fundamental limitation of ex vivo therapies, persistence, and thereby unlock waves of potentially disruptive therapies across a variety of cell types.

Safety Switch for Hypoimmune Cells

We are actively investigating approaches to control hypoimmune cells after administration into the patient. If necessary, the aim of these “safety switches” would be to provide a mechanism to eliminate hypoimmune cells within the body in a targeted fashion, in scenarios where the cells are not in a location where physical removal is viable. Such a safety switch would be beneficial to mitigate the potential risk of a hypoimmune cell becoming infected with a virus or undergoing oncogenic transformation, in light of the immune evasion modifications to these cells.

One approach we are exploring as a safety switch is to re-sensitize the hypoimmune cells to innate cell killing via administration of a blocking anti-CD47 antibody. We have tested the effectiveness of this approach in iPSCs and teratomas (a particular tumor formed by pluripotent cells with histological features from all three germ layers), both bearing the hypoimmune modifications. Using hypoimmune NHP iPSCs we observed in vitro that the addition of an anti-CD47 antibody binds to and blocks CD47 expressed in the hypoimmune cells and restores the sensitivity to the missing-self killing response mediated by NK cells. We also assessed this strategy in mouse experiments, where the animals were transplanted in with human iPSCs which then formed small teratomas. We have shown that treatment with an anti-CD47 antibody resulted in the loss of immune evasion and the rapid killing of these transplanted cells. We have identified several additional safety switches with in vivo activity and intend to continue to explore them, potentially including multiple safety switches in therapeutic programs moving forward.

Anti-CD47 Administration Results in the Rapid Clearance of Hypoimmune NHP iPSCs in vitro

On left: Hypoimmune NHP iPSCs do not induce killing by NK cells in an in vitro killing assay; On right: by contrast, Anti-CD47 antibody treated hypoimmune NHP iPSCs are no longer able to evade missing-self responses mediated by NK cells and are killed rapidly.

Anti-CD47 Administration Results in the Rapid Clearance of Human iPSCs derived Teratomas in a Humanized Mouse Model

On left: Human iPSC proliferate (as visualized by luminescence of live cells) and form teratoma in NSG mice (n=3) with adoptive transferred human NK cells. Administration of isotype control has no impact on HIP survival. On right: Blocking of CD47 in vivo results in killing of HIP iPSCs (as visualized by luminescence of live cells) in NSG mice (n=5) with adoptive transferred human NK cells.

CD47 overexpression is differentiated in inhibiting “missing self” response relative to other approaches

As part of our ongoing program to further refine our hypoimmune technology we evaluated the effectiveness of the overexpression of CD47 in comparison to other molecules that have at least some ability to inhibit innate immune responses. We carried out these head-to-head comparisons in K562 cells, a naturally MHC class I and class II deficient cell line. The lack of the MHC class I molecule should result in virtually instantaneous cell killing by stimulated innate immune cells such as NK cells due to the activation of the “missing self” response. We compared three molecules, HLA-E, HLA-G and PDL-1, that have previously been proposed to have a role in inhibiting innate immune responses versus CD47. In this assay, overexpression of these three molecules conferred limited protection from NK cell killing in contrast to CD47 overexpression. This difference in activity may be the result of the more ubiquitous presence of the receptor for CD47 receptor on innate immune cells relative to the presence of receptors for these other immunomodulators. While these results do not rule out a role for these other molecules in inhibiting NK cell responses, they suggest that CD47 may be sufficient to nullify the NK cell-mediated missing-self response.

Panels above show in vitro killing assays mediated by NK cells. Cells missing MHC molecules are killed by NK cells, as measured by rapid decline in cell index. Overexpression of immunomodulatory molecules such as HLA-E, HLA-G or PDL-1 in cells missing MHC molecules did not block NK cell killing. By contrast, overexpression of CD47 blocked NK cell mediated “missing-self” response.

Our ex vivo Cell Engineering Pipeline

Allogeneic T Cell Program (SC291, SC276, SC255)

Our allogeneic T cell program utilizes T cells from healthy donors to generate CAR T therapies that will initially target CD19, a protein expressed on the cell surface of B cell malignancies, to treat patients with refractory lymphoma. We believe that applying the hypoimmune technology to allogeneic T cells gives us an opportunity to create differentiated allogeneic CAR T therapies.

We believe our allogeneic T cell and T cell fusosome discovery programs provide us with two potentially disruptive programs to address the limitations of adoptive T cell therapy for cancer, each with idiosyncratic risks and opportunities. We also believe each approach can address separate and valuable opportunities if they are both successful. Specifically, our allogeneic T cell program offers the opportunity to perform multiple gene edits in a T cell, which may allow us to make intentional modifications to control T cell function or to deliver more complex chimeric receptors and signal integration machinery to enable the T cell to distinguish tumor cells based on surface antigen combinations to improve the specificity of targeting. These approaches may prove especially valuable in

targeting solid tumors, which have remained largely refractory to CAR T approaches to date. We also have an earlier-stage program looking to differentiate hypoimmune iPSCs into T cells. While we are still working to successfully create the appropriate T cells from an iPSC, we expect that progress with our allogeneic T cell program will also inform the iPSC T cell program. Separately, the fusogen technology allows for the in vivo generation of CAR T cells in a patient, offering a distinct advantage in terms of manufacturability and scalability that may enable the introduction of gene-modified T cells earlier in the course of a patient’s therapy. Additionally, modifying the T cells inside the body without the need for ex vivo manipulation of the cells may generate CAR T cells with more favorable attributes.

We intend to develop our CD19 allogeneic T cell therapies with the goal of submitting an IND for SC291 in 2022. We intend to follow this with SC276, a CD22-targeting allogeneic CAR T which could be combined with targeting CD19 and offers the potential benefit of higher and more durable complete response rates, with an IND submission as early as 2023. We are also advancing our SC255 allogeneic T cell program targeting BCMA for multiple myeloma, with the goal of submitting an IND in the next several years.

Background on B cell Malignancies

B cell malignancies represent a spectrum of cancers including non-Hodgkin Lymphoma (NHL), chronic lymphocytic leukemia (CLL), acute lymphoblastic leukemia (ALL), and multiple myeloma (MM) and result in over 100,000 deaths per year in the United States and Europe. See the subsection titled “—in vivo Cell Engineering Pipeline—Background on B Cell Malignancies” for further background discussion. In addition to our in vivo cell engineering technology, we believe our ex vivo cell engineering technology also has the potential to address B cell malignancies.

Current Treatment Landscape and Unmet Need

We believe our hypoimmune edited cells have the potential to create a differentiated platform for developing allogeneic T cells. There are two major hurdles to the use of allogeneic T cells. The first is the risk of graft versus host disease, in which the allogenic donor T cells target and kill recipient tissues. Multiple CAR T cell product candidates in clinical development have managed to prevent this reaction through gene edits targeting components of the T cell receptor such as TCR-alpha gene. The more significant challenge has been HvGR, in which the patient’s immune system kills the transplanted T cells. One strategy to approach this challenge has been to essentially eliminate the patient’s immune system, neutering its ability to find and destroy the transplanted allogeneic CAR T cells. This strategy has two limitations. First, the patient is at risk of severe infections during this period of substantial immune suppression. Second, as the immune system returns, it will inevitably reject the allogeneic CAR T cells, limiting the duration that these therapeutic cells are in the body. Experience with autologous CAR T cells in patients with B cell malignancies has demonstrated that persistence of CAR T cells is important for the durability of response. Thus, the ability to effectively prevent long term rejection of an allogeneic CAR T therapy without significant immune suppression would be a major advance. We are aware of other efforts to develop allogeneic CAR T cell products that focus on overcoming the adaptive immune system (T and B cells). Our technology addresses rejection mediated by both the adaptive and innate immune systems, giving us the potential to create a differentiated allogeneic CAR T solution.

Our Allogeneic T Cell Program Approach

Our hypoimmune technology is designed to hide the cell from the patient’s immune system, and we are applying this technology to manufacture allogeneic CAR T cells. We intend to utilize T cells from healthy donors into which we will introduce the CAR gene and make the gene modifications necessary to overcome graft versus host disease and to incorporate our hypoimmune technology in an effort to address host versus graft response. We then intend to expand these cells ex vivo, with a goal of making many batches from a single donor as well as creating comparable CAR T cells from various healthy donors. These allogeneic CAR T therapies could be frozen and delivered as an “off the shelf” product for cancer patients without the need for severe immunosuppression.

Preclinical Data

For our preclinical studies, human donor-derived T cells were genetically modified ex vivo, to generate cells bearing the hypoimmune edits (disruption of MHC class I/class II; overexpression of CD47), TCR-alpha disruption (to mitigate graft versus host disease) and the expression of a CD19 CAR. These cells were then tested in vivo for their tumor-killing activity in a human xenograft mouse model for leukemia (Nalm-6). These preclinical data suggest that the hypoimmune edits do not interfere with CAR T killing activity. We observed clearance of the leukemic cells by the hypoimmune CD19 CAR T cells and the potency of these cells was comparable to unmodified CD19 CAR T cells, which are similar to CAR T cells currently in clinical use.

Hypoimmune Donor-Derived CD19 CAR T Cells Clear Leukemia Cells in a Human Xenograft Mouse Model at Levels Comparable to Unmodified CD19 CAR T Cells

Activity of hypoimmune donor-derived CD19 CAR T in a mouse leukemia xenograft model (Nalm-6). Note that when compared to untreated controls, infusion of unmodified CD19 CAR T or hypoimmune CAR T results in eradication of leukemia cells. Both cohorts of CAR T treated mice had significantly reduced tumor burden when compared to control as early as D7 (p≤0.0001; One-way ANOVA Bonferroni) with no significant difference between either of the treatment arms.

Furthermore, the hypoimmune CD19 CAR T cells were protected from immune system rejection in humanized mice, and no evidence was observed of either adaptive or innate immune system activation, in contrast to the unmodified CART cells:

Absence of T Cell, B Cell, NK Cell, and Macrophage Responses in a Human Xenograft Mouse Model Following Injection of Hypoimmune Donor-Derived CD19 CAR T Cells

Immune cells from humanized animals receiving hypoimmune donor-derived CD19 CAR T showed no response when exposed to hypoimmune CAR T cells in vitro. Delivery of hypoimmune cells did not result in production of donor-specific antibodies, as evidenced by binding of IgM antibodies to the surface of donor cells (increased mean fluorescent intensity, MFI). Hypoimmune-edited cells were not susceptible to killing by NK cells nor macrophages, indicating protection from the “missing self” signal.

Development Plan and Key Next Steps

Process development work is ongoing within the Technical Operations team to develop scalable manufacturing processes to generate quality and consistent allogeneic T cell product candidates. In parallel, our cell engineering team is developing an efficient and specific gene editing platform to enable manufacturing of our allogeneic T cell product candidates.

The next major milestones are to complete Good Laboratory Practice (GLP) studies and GMP manufacturing with the goal of submitting an IND for SC291 in 2022. We believe that early data from the SC291 study will help us understand the therapeutic potential of this therapy, as the ability to evade immune detection with corresponding enhanced persistence of the CAR T may predict higher and more durable complete responses for patients. Additionally, it will give us insight into the potential of this hypoimmune platform for additional CAR T programs targeting other antigens and cancers as well as the potential for the platform more broadly in areas beyond cancer. Following the IND submission for SC291, we are progressing SC276, a CD22-targeting allogeneic CAR T which could be combined with targeting CD19 and offers the potential benefit of higher and more durable complete response rates, with the goal of submitting an IND as early as 2023. We are also advancing our SC255 allogeneic T cell program targeting BCMA for multiple myeloma, with the goal of submitting an IND in the next several years.

Beta Cell Program

Our beta cell program aims to restore lifelong glucose control in Type I diabetes mellitus (T1DM), patients by transplanting hypoimmune iPSC-derived beta cells. Current therapies for T1DM require continual management, and we believe that effectively restoring beta cell functionality will meaningfully improve patient outcomes for patients with T1DM. We intend to develop this program with the goal of submitting an IND for SC451 as early as 2023.

Background on Type 1 Diabetes

T1DM is an autoimmune disease in which the patient’s immune system destroys its own pancreatic beta cells. The destruction of these cells leads to complete loss of insulin production and a metabolic disease wherein patients are unable to control their blood glucose levels. Often called “juvenile diabetes”, this disease commonly has its onset in adolescence. Beta cells reside in specialized hormone-producing clusters within the pancreas called the Islets of Langerhans. In T1DM, activated T lymphocytes infiltrate the islets and selectively kill the beta cells, progressively reducing the body’s capacity to produce insulin. Once the reserve capacity of beta cells is exhausted, blood glucose rises, and the patient will have a life-long battle to control blood glucose levels.

T1DM affects 1.6 million adults in the United States, and there are approximately 20,000 new cases diagnosed per year in patients under the age of 20. In Europe there are an estimated 2.4 million adults with T1DM, and 300,000 under the age 20, with 31,000 new cases of T1DM diagnosed each year. Combining prevalence in the United States and Europe yields a pool of approximately 4 million patients with T1DM.

Current Treatment Landscape and Unmet Need

Insulin injection is the main treatment option for T1DM. Despite significant advances in types of insulins, glucose monitoring, and insulin pumps, life expectancy for T1DM is still approximately 15 years shorter than for people without diabetes. Patients are at risk from acute complications of hyperglycemia, including diabetic ketoacidosis and coma. Conversely, they are also at risk of hypoglycemic episodes, particularly at night, which can lead to the “dead in bed” syndrome, thought to result from cardiac arrhythmias induced by low glucose. Long term elevations in blood glucose levels have particularly devastating effects on arteries and capillaries, resulting in premature myocardial infarction, stroke, limb ischemia, gangrene, kidney failure, and blindness due to diabetic retinopathy. “Insulin pumps,” which feature a computerized system for sensing blood glucose and delivering appropriate doses of insulin, have improved glycemic control. Notably, data from the FDA indicate that issues with insulin pumps are among the most frequently reported problems in their database. All current therapies require patients to carefully monitor their dietary intake, which, while inconvenient in adults, is a frequent point of failure in adolescents.

Pancreas transplantation for uncontrollable diabetes was first performed in the 1960s, and this established the principle that replacing the beta cells (here in the context of the whole pancreas) could restore physiological glucose control. Pancreas transplants are complicated surgical interventions, require lifelong immunosuppression, and are limited due to organ availability. Nevertheless, some 30,000 pancreas transplants have been performed worldwide to date.

Because of these challenges, the biomedical community began exploring pancreatic islet transplantation in the 1970s. This process involves enzymatic digestion of a donor pancreas and isolation of the Islets of Langerhans followed by delivery of these cells to an appropriate site in the body where the islets can engraft and become well vascularized. The major lessons from islet transplantation have been that glucose homeostasis can be restored, insulin-independence can be achieved, hemoglobin A1C levels (a marker of long-term glucose levels) can be normalized, and severe episodes of hypoglycemia can be reduced. As with an organ transplant, patients must be immune suppressed to prevent immune rejection of the transplanted cells. In addition to complications from this immune suppression and the lack of cell availability, the principal limitation of islet transplantation has been the therapy’s durability. Most patients lose glucose control over months to years and eventually become insulin-dependent again, primarily due to immune rejection of the allogeneic islets.

Our Beta Cell Program Approach

The goal of our beta cell hypoimmune program is to restore lifelong glucose control in T1DM patients by transplanting hypoimmune iPSC-derived beta cells, including beta cells. Our goal is to create a therapy that restores the body’s normal beta cell mass, giving patients physiologically appropriate glucose sensing ability and insulin secretion. We believe this therapy could reduce, or even eliminate, the hypoglycemia and hyperglycemia in diabetic patients, potentially enabling less onerous and costly treatment, fewer complications, and longer life expectancy, resulting in a meaningfully improved quality of life.

We focus our efforts around three goals: (i) deriving highly functional beta cells from PSCs, (ii) genetically modifying these cells to evade allogeneic immune responses, and (iii) genetically modifying these cells to evade autoimmune destruction of beta cells. This strategy requires building on lessons from pancreatic islet transplantation, recent advances in understanding pancreatic islet developmental biology, and our hypoimmune technology.

Deriving beta cells from iPSCs has the potential to solve limitations associated with donor pancreas and improve the overall product quality and product consistency. iPSCs have the potential to create a virtually limitless supply of these cells. Our program uses proprietary differentiation protocols to generate mature beta cells with glucose control comparable to primary human islets, as evidenced by our animal studies. Finally, we intend to modify the genomes of the iPSCs in order to apply our hypoimmune technology. If successful, the hypoimmune gene modifications will protect these cells from both auto-immune and allogeneic rejection

by the patient’s immune system and potentially remove the need for toxic immunosuppression in transplant recipients. Hypoimmunity also should eliminate the need for physical separation of the beta cells from the rest of the body by a device or encapsulation technology, which may allow for tighter control of glucose by eliminating the lag time between glucose sensing and insulin secretion.

Preclinical Data

We are developing a proprietary protocol that will differentiate hypoimmune iPSCs into mature, glucose-sensitive, insulin-secreting beta cells based on licensed technology from Washington University in St. Louis. This technology enables differentiation of beta cells at a greater purity and with superior function compared to published stem cell-based protocols. The principal function of beta cells is to maintain steady levels of glucose in circulation. The beta cells sense when glucose levels rise in the bloodstream and release insulin in response. In vitro, our beta cells respond to glucose and robustly secrete insulin at an equivalent level to primary human islets, as depicted in the figure below.

Human iPSC-Derived Beta Cells Exhibit Glucose-Induced Insulin Release

Human islets from cadaveric pancreases (gold standard) exhibit robust insulin secretion in response to an increase in glucose levels. Human iPSC-derived beta cells using technology licensed from Washington University in St. Louis show similar dynamics of insulin secretion to the cadaveric islets.

These stem cell-derived beta cells were tested in a mouse model of Type I diabetes induced by the beta cell toxin, streptozotocin. When transplanted into the kidney of the diabetic mice, these beta cells normalize glucose levels in an equivalent fashion to primary human islets. The diabetic glucose levels return when the grafts are surgically excised (nephrectomy). Similar to the human phenotype, diabetic mice cannot normalize circulating glucose levels following a glucose injection. Following transplantation of our beta cells, these mice rapidly normalized blood glucose in an equivalent fashion to both non-diabetic mice and diabetic mice that received human primary islet transplants.

In vivo Performance of iPSC-Derived Beta Cells in a Mouse Model of T1DM

Top: Normalization of blood glucose levels after transplantation of cadaveric human islet cells or iPSC-derived islet cells obtained by planar or suspension differentiation (Washington University technology). Note the rapid normalization of blood glucose with cadaveric and stem cell-derived islets with the planar protocol, with slower normalization using the suspension protocol. In all groups, removal of the graft by nephrectomy re-induced diabetes, indicating the correction resulted from the transplant. STZ, streptozotocin, is a toxin for beta-islet cells that induces diabetes in animal models. Bottom: Normalization of blood glucose after glucose injection by transplantation of cadaveric islet cells or iPSC-derived islet cells. Note the more complete normalization using the planar protocol. Groups are defined by the same symbols shown in the middle panel. From Hogrebe et al, Nature Biotechnology 2020.

We ran an experiment to better understand whether hypoimmune edits impair the function of islet cells and to confirm that these edits enable the islets to evade immune responses. For these experiments, we made the hypoimmune genetic modifications in human islets isolated from cadaveric human donors. These cells were transplanted intramuscularly, without immunosuppression, into a humanized mouse strain, a mouse strain engrafted with human immune cells that mimic certain aspects of the human immune system. Type 1 diabetes was induced in this mouse model with streptozotocin, and the ability of transplanting islet cells to restore normal glucose levels while evading immune rejection was tested. We found that hypoimmune edited human donor islets were successful in restoring glucose control while unmodified islets failed to do so. The key driver of this difference was that, while human donor islets with hypoimmune edits survived for the full duration of the experiment (approximately 30 days), unmodified islets were rapidly rejected within a week. These data demonstrate that hypoimmune edits do not impair the ability of cadaveric islet cells to restore glucose homeostasis, but they do prevent rejection of these cells in mice with humanized immune systems.

We next tested whether our hypoimmune edited cells could survive in serum containing autoreactive antibodies and T cells from patients with Type I diabetes. We exposed pancreatic islets with or without our hypoimmune edits to serum and T cells from patients with Type I diabetes. As expected, both T cells and antibodies from patients recognize and kill unedited pancreatic islets. In contrast, these T cells and antibodies from Type I diabetics did not recognize or kill our hypoimmune edited islet cells. The results from this experiment increase our confidence that, in addition to preventing allogeneic rejection, our hypoimmune technology may allow islet cells to survive autoimmune killing in patients with Type I diabetes without the need for immunosuppression.

Hypoimmune human donor pancreatic islet cells evade immune detection and regulate glucose levels

Unmodified human donor derived pancreatic islets (Row 1) are rapidly killed by PBMCs (immune cells) from Type 1 Diabetic donors but are not recognized by immune cells from healthy volunteers. Human donor derived pancreatic islets bearing hypoimmune modifications (Row 2) are not recognized by the PBMCs of healthy volunteers or type 1 diabetics

Left graph: Unmodified human donor derived pancreatic islets (unmodified islet cells) are readily recognized by antibodies in the serum of Type 1 Diabetic patients. Antibody binding is corelated with higher values of Maximum Fluorescent intensity (MFI) Right graph: Human donor derived pancreatic islets bearing hypoimmune modifications are not recognized by antibodies in serum of healthy volunteers or Type 1 diabetic patients

Hypoimmune human donor pancreatic islet cells survive in humanized mice and regulate glucose levels

Top image panel on left: Unmodified human donor derived pancreatic islets injected intramuscularly into humanized mice are rejected by recipient mice by D9. Viable cells are marked via a bioluminescent marker that is lost when transplanted cells are no longer viable. Top image panel on right: By contrast hypoimmune human donor derived pancreatic islet cells survive in humanized recipient mice till D29 when the experiment was terminated

Lower graphs are serum glucose measurements in the humanized mice. Diabetes was induced via streptozotocin treatment that results in loss of glycemic control as visualized by an immediate and sustained increase in serum glucose. Transplantation of unmodified human donor derived pancreatic islets had no effect on serum glucose levels. By contrast hypoimmune human donor derived islets successfully lowered serum glucose levels as measured by a glucose tolerance test.

Development Plan and Key Next Steps

Our work is currently focused on manufacturing GMP-grade, gene-edited, pluripotent stem cell banks; scaling manufacturing; and characterizing the product. We are working through the process development and IND-enabling studies to allow for an IND submission for SC451 as early as 2023.

GPC Program

Our GPC program aims to deliver healthy allogeneic GPCs, the precursors to both astroglia and myelin-producing oligodendrocytes. This program has the potential to treat myelin and glial-based disorders, which represent a broad group of debilitating neurological disorders, such as multiple sclerosis and a number of neurodegenerative disorders, none of which have effective treatment alternatives. We intend to develop our stem cell derived GPC therapies for secondary progressive multiple sclerosis, Pelizaeus-Merzbacher disease other disorders of myelin, Huntington’s disease, and other astrocytic diseases. Our goal is to submit three INDs for SC379 in the next several years.

Background on Myelin and Glial Based Disorders

Glial cells are the support cells of the human CNS. The two major types of CNS-derived glial cells are oligodendrocytes—the cells that produce myelin, the insulating substance of the brain’s white matter that enables neural conduction and astrocytes, the support cells of neurons and their synapses. These two kinds of glial cells arise from human GPCs (hGPCs), are responsible for remyelination in the injured and demyelinated adult brain and spinal cord.

Diseases of glial cells are among the most prevalent and disabling conditions in neurology. These disorders include the disorders of oligodendrocyte loss and myelin failure and the disorders of astrocytes, which include a number of neurodegenerative and psychiatric disorders. What all these disorders have in common is a significant glial contribution to their pathogenesis, and a lack of disease-modifying treatment options.

Congenital Leukodystrophies. A number of hereditary disorders of oligodendrocyte loss or dysfunction are characterized by a failure in myelin synthesis or structural stability. Tens of thousands of children in the United States suffer from diseases of myelin loss. The most prototypic example of this class of diseases is Pelizaeus-Merzbacher disease (PMD), an X-linked leukodystrophy most often manifesting in male infants and young boys, caused by mutations in the oligodendrocytic PLP1 gene, which results in widespread hypomyelination. There is no treatment for PMD, which is typically fatal in childhood. We intend to deliver intracerebral transplants of stem cell-derived GPCs to the brains of PMD patients, with the goal of replacing PLP1 mutant oligodendrocytes with healthy cells capable of producing normally compact myelin. Prevalence of PMD in the general population is estimated to be approximately 1 in 100,000 in the United States. While we are initially targeting PMD as our proof of concept, congenital leukodystrophies as a group affect a more significant population, or about 1 in 7,600 births.

Multiple Sclerosis (MS). MS is a debilitating disease characterized by both inflammatory myelinolysis and degenerative axonal loss. There are two major forms, the initial relapsing remitting form, known as RRMS, and its later progressive neurodegenerative phase designated secondary progressive MS (SPMS). RRMS is characterized by clearly defined attacks with new or increasing neurologic symptoms. In contrast, SPMS is characterized by progressive neurodegeneration with a loss of neurons, including those that were previously demyelinated during the RRMS phase of the disease. The demyelination occurs in a diffuse fashion throughout the adult brain and appears to reflect a loss of axonal support by local oligodendrocytes. The delivery of GPCs into such chronically demyelinated brain may offer tangible benefits through the oligodendrocytic engagement of axons, as well as by myelin repair. MS is highly prevalent, with estimates of up to 1.0 million in the United States, 600,000 in Europe, and 2.8 million patients globally. Approximately 85% of MS patients receive a diagnosis of RRMS initially while 15% of patients are diagnosed with primary progressive MS (PPMS). Up to a third of RRMS patients transition to secondary progressive MS within a decade if untreated, and most will progress to SPMS within 20-25 years of diagnosis. Success with a stem cell derived GPC product in SPMS, and especially with a hypoimmune product, could enable further expansion into the RRMS patient population.

Huntington’s Disease (HD). HD is a neurodegenerative disorder in which glial pathology appears to make a significant causal contribution. It is an autosomal dominant disorder characterized by abnormally long CAG repeat expansions in the first exon of the Huntingtin gene. The encoded polyglutamine expansions of mutant huntingtin protein disrupts its normal functions and protein-protein interactions, ultimately yielding widespread neuropathology, most rapidly evident in the neostriatum. We have found that glial pathology is a major contributor to the functional deficits of HD and repairing the glial pathology has significant and positive effects in animal models. There are approximately 41,000 symptomatic Americans and more than 200,000 at-risk of inheriting HD. In Europe, there are approximately 50,000 patients with HD.

Current Treatment Landscape and Unmet Need

Congenital Leukodystrophies. There are no viable treatment options for these conditions, only supportive and palliative therapies for symptoms as they present.

MS. Current treatments for MS are largely limited to treatments for RRMS; few treatments are approved for SPMS, and these have at best marginal efficacy in delaying disease progression; none are restorative. Currently approved treatments for RRMS may be divided into three broad categories of disease modifying therapies: (i) first line injectables (such as beta-interferons, Copaxone), (ii) newer oral agents (such as Tecfidera, Gilenya, Mayzent, Zeposia), and (iii) high-efficacy agents (such as Tysabri, Lemtrada, Ocrevus). Despite many recently successful drug launches in the RRMS space, these drugs still only slow the progression of disease and aid in the recovery from attacks, and there remains no treatment that confers functional restoration or effective cure for this disease.

HD. Currently, there is no treatment to stop or reverse Huntington’s disease. Treatment is limited to several medications that can help minimize symptoms, including the drug tetrabenazine, antipsychotic drugs, antidepressants, and tranquilizers.

Our GPC Program Approach

Our approach to treat myelin and neurodegenerative disorders is via the delivery of healthy allogeneic stem cell-derived GPCs. We have developed methods for producing and isolating these cells from pluripotent stem cells and delivering them in the purity and quantities necessary for their replacement of endogenous diseased cells. We believe that both the myelin disorders and glial-based neurodegenerative conditions have compelling potential for our ex vivo therapy.

Preclinical Data

Congenital Leukodystrophies. The capacity of stem cell-derived GPCs for remyelination has been conducted in animal models of congenital hypomyelination. Our collaborators used newborn shiverer mice that have a genetic defect in myelin basic protein (MBP), resulting in their neurons being hypomyelinated and the mice having a shortened lifespan. When iPSC-derived hGPCs were transplanted into these mice, the cells spread widely throughout the brain, developing as astrocytes and oligodendrocytes. These oligodendrocytes generated mature myelin that effectively restored neuronal conductance and prolonged survival in the transplanted

mice. We believe that these data suggest the feasibility of iPSC-derived hGPC implantation in treating childhood disorders of myelin formation and maintenance, as depicted in the figure below:

hGPCs myelinate widely to greatly extend the survival of hypomyelinated mice. A, Dot map indicating distribution of human iPSC-derived GPCs at 7 months of age, following neonatal engraftment in a shiverer mouse brain. Widespread colonization and chimerization of the host brains by iPSC-derived hGPCs is evident (human nuclear antigen, red). B, iPSC-derived hGPC-derived myelination in shiverer forebrain, at 7 months; section 1 mm lateral to A. Myelin basic protein (MBP)-immunoreactivity (green) is all human donor-derived. C, D. Myelination in sagittal sections taken at different mediolateral levels from 2 additional 7 month-old mice, each engrafted with iPSC-derived hGPCs at birth. E, Kaplan-Meier plot of survival of iPSC-OPC implanted (n=22) vs. saline-injected (n=19) control mice. Scale: A-B, 2 mm. Adapted from Wang, Cell SC 2013.

MS. Our prior studies established the ability of stem-cell derived hGPCs to myelinate the developing shiverer brain and rescue the afflicted mice; however, the experimental subjects were neonates, not adults. Until recently, it was unclear whether GPCs are able to migrate extensively in adult brain tissue, as would be required for the repair of diffusely demyelinated adult brains. To explore whether the introduction of stem-cell derived hGPCs delivered directly into the adult brain could remyelinate axons in the setting as might be encountered clinically in multiple sclerosis our collaborators studied three different biologic models. First, it was shown that stem-cell derived hGPCs can disperse within and myelinate the brains of adult shiverer mice (as depicted in the figure below). Second, it was shown that neonatally-engrafted hGPCs engrafted as a neonate can generate new oligodendrocytes and remyelinate demyelinated axons after chemically induced demyelination. This result demonstrated the ability of already-resident hGPCs to remyelinate previously myelinated axons after a new demyelinating insult as an adult, as well as the ability of transplanted hGPCs to reside as a functional reservoir of new myelinogenic cells in the host brains. Third, it was shown that hGPCs transplanted into the adult brain after chemically induced demyelination can remyelinate denuded axons. These data suggest that transplanted hGPCs can disperse broadly and differentiate as myelinogenic cells in the adult brain, and that they are able to remyelinate demyelinated axons and white matter lesions of the brain after an insult as an adult.

hGPCs Mediate Robust Myelination After Transplantation into the Adult Shiverer Brain

Human GPCs proved both highly migratory and robustly myelinogenic, after delivery to the hypomyelinated adult shiverer x rag2-/- brain. A, By 19-20 weeks of age (mice were injected as post-weaning adults, at 4-6 wks) the injected cells had dispersed broadly throughout the forebrain white matter. B, hGPCs delivered to myelin wild-type rag2-/- mice distributed throughout both gray and white matter. C, Oligodendrocyte differentiation and myelinogenesis by donor hGPCs was robust, with myelination of brain regions that would typically be demyelinated in shiverer mice D, a higher power image of C shows the high proportion of donor cells in those brain regions. Note that DAPI marks all nuclei, hN marks the hGPCs, and MBP marks the remyelinated regions in C and D. From Windrem et al, Cell Reports 2020.

HD. Our collaborators explored the cellular basis for HD related glial pathology and identified significant defects in potassium channel and glutamate uptake mechanisms in HD glia, which appeared to account for both the glial pathology and its deleterious effects on synaptic function. Together, these studies suggest a critical role for glial pathology in the progression of HD and suggested the potential for glial cell replacement as a therapeutic strategy in HD, and more broadly, to other neurodegenerative diseases in which glial pathology might be causally contributory. It was confirmed in preclinical mouse studies that stem-cell derived hGPC transplant ameliorated both the neuronal and glial pathology of HD by restoring synaptic homeostasis and normal synaptic function to the most affected regions of the host brain.

The majority of the studies with human GPCs thus far have been xenogeneic grafts of human GPCs to neonatal or adult mice or rats (and in a small sample POC study limited to adult tissue-derived hGPCs, NHPs). Our collaborators have also performed studies with murine GPCs transplanted into both developing and adult mice, which have confirmed allogeneic GPC migration and integration. However, we have no assurance that human GPC engraftment of human brain will result in the widespread migration and colonization of host brain that is seen with xenogeneic grafts. To better model the human-to-human graft paradigm, our collaborators have established a new model to evaluate if GPC engraftment will result in migration and colonization in a host brain. This model allows observation of the competitive interactions of the two separately tagged human GPC populations. The human-into-human grafts expanded and integrated well in their humanized host, with competitive interactions. As might be anticipated in the clinical setting of healthy cells being transplanted for the purpose of replacing lost or diseased hGPCs, the healthy donor cells outcompete both diseased and older cells to ultimately colonize the hosts. These data have provided preclinical assurance of the fundamental premise of our approach, that healthy human donor cells can replace lost or diseased human cells in vivo. That said, this determination remains to be made in patients.

GMP Grade Stem Cell Derived hGPCs for Clinical Studies

A protocol to direct differentiation of human ESCs, as well as iPSCs, to hGPCs has been established. These hGPCs cells remain bipotential for astrocytes and oligodendrocytes, and they differentiate to either fate depending on local signaling.

This protocol has been transferred to a GMP facility in order to enable production of clinical grade cells for both safety and efficacy testing. These cells have been validated to robustly remyelinate shiverer mouse brains upon intracerebral transplantation. We plan to use these cells for our IND-enabling studies and initial clinical trial material.

Development Plan and Key Next Steps

Progression of SC379 to IND is planned to follow completion of definitive safety and toxicology studies. Definitive preclinical efficacy studies using the anticipated clinical product are also planned and will replicate studies that we have published. Since GPCs are not a terminally differentiated cell type and divide and differentiate in vivo post-transplantation, we will continue to assess potential safety risks, including the risk of tumorigenicity. We expect to submit IND applications for SC379 for SPMS, PMD, and HD in the next several years.

Cardiomyocyte Program

Background on Heart Failure

Heart failure (HF) is a classic example of a disease of cell loss, ideally suited to the application of ex vivo engineered cells. The clear but ambitious goal of our program is to replace missing cells after a myocardial infarction, commonly known as a heart attack, in an attempt to restore heart function and improve outcomes for patients. HF is a life-threatening syndrome, and patients with HF have a mortality rate of 20-30% within one year of diagnosis and a mortality rate of around 50% within five years of diagnosis.

HF with reduced ejection fraction (HFrEF), is a severe form of HF where heart muscle is unable to contract, and therefore pump, adequately. HFrEF is most frequently a consequence of a loss of heart muscle cells (cardiomyocytes), following a myocardial infarction. In the United States, there were approximately 380,000 deaths associated with HF in 2018 according to the United States Centers for Disease Control and an overall prevalence of approximately 6 million people with HF, with similar numbers in Europe.

In general, HF has been a challenging area for drug and device development, including only one new drug, Entresto, approved in the last 20 years, and a limited number of devices introduced including electrical resynchronization therapy and implantation of left ventricular assistance devices (LVADs). These approaches provide only symptomatic relief and do not address the underlying loss of cardiomyocytes associated with HFrEF. As a result, HFrEF currently remains a progressive and deadly disease with a large unmet need worldwide.

To date, efforts to develop cell-based therapies to address this unmet need have provided little evidence of clinical benefit. Importantly, these attempts have typically utilized cells such as bone marrow-derived mononuclear cells and mesenchymal stromal cells where any potential benefit would be limited to paracrine mechanisms and not the direct replacement of lost cardiomyocytes.

Our cardiomyocyte program aims to directly regenerate the heart, by replacing lost cardiomyocytes with iPSC-derived cardiomyocytes, with the goal of restoring heart muscle and increasing ejection fraction, which is the percentage of blood the heart pumps with each heartbeat. Replacement of lost cardiomyocytes with iPSC-derived cardiomyocytes that engraft and function correctly has the potential to prevent or even reverse the progression of HFrEF.

Developing an ideal stem cell-derived cardiomyocyte therapy involves many steps, including:

Differentiating Cardiomyocytes at Scale that Engraft, Beat Correctly and in Synchrony with the Host Heart Muscle, and Improve Host Heart Function

Scientists, including Dr. Chuck Murry, our Senior Vice President, Head of Cardiometabolic Cell Therapy, have been working for over 20 years towards developing a regenerative therapy for HFrEF with the goal of transplanting cardiomyocytes derived from

human pluripotent stem cells that engraft, function, and persist inthe human heart in vivo. The groundwork for potential future clinical development has been laid by key breakthroughs such as the ability to direct the differentiation of stem cells selectively into cardiomyocytes, including producing pharmaceutical grade cardiomyocytes at large scale in bioreactors, and the ability to transplant such cardiomyocytes to induce remuscularization of injured hearts.

Initial preclinical attempts to remuscularize the infarcted heart were unsuccessful due to death of the transplanted cells within a few days of delivery. None of the animals with failed engraftment showed improvement in cardiac function, indicating that engraftment is essential for functional improvement. Our collaborators developed a pro-survival cocktail that kept cells alive through the rigors of transplantation, allowing the cardiomyocytes to self-assemble into new muscle tissue and induce ingrowth of new blood vessels and connective tissue from the surrounding heart muscle. Once engraftment was successful, cardiac function improved. As our collaborators’ capabilities to scale cell manufacturing increased, studies progressed from mice to rats to guinea pigs, all showing improved function:

Human ESC-Cardiomyocytes Improve Function in Injured Rat and Guinea Pig Hearts

Left panel: functional rescue in rat. All groups showed comparably reduced fractional shortening after infarction at 2 days pre-transplantation (Pre-Tx). At 28 days post-transplantation there was preservation of fractional shortening in animals receiving hESC-cardiomyocytes, with deterioration of function in all other groups. PSC, pro-survival cocktail. SFM, serum-free media. **p<0.01; # p<0.05 vs. 2 days Pre-Tx. Right panel: functional rescue in guinea pig. Following cardiac injury at 2 days before transplantation, all groups showed comparably reduced fractional shortening. At 28 days post-transplantation there was preservation of function in animals receiving hESC-cardiomyocytes (hESC-CM), with deterioration of function in other groups. *p<0.05. **p<0.01. †p<0.05 vs 2 days Pre-Tx. From Laflamme et al, Nature Biotechnology 2007 (left) and Shiba et al, Nature 2012 (right).

Current methods demonstrate regeneration of the hearts of large animals including pigs and NHPs by transplanting human ESC-derived cardiomyocytes (hESC-CM). The figure below shows low magnification microscopic images from NHP hearts that were infarcted and then received either hESC-CM or saline controls. The replacement of heart muscle by scar tissue is evident in the saline-treated heart, whereas human heart muscle has repopulated the infarct in the hESC-CM treated group.

Remuscularization of the Heart of an NHP by hESC-CM Transplantation

The saline-treated heart (left) shows infarct scar tissue (blue) replacing the myocardium lost to infarction. The hESC-CM treated heart (right) shows a large graft of human heart muscle (green) replacing the myocardium lost to infarction. From Liu et al, Nature Biotechnology 2018. Scale bar, 5 mm.

We conducted an experiment to help us understand the mechanism of action and learn whether the transplanted muscle beat in synchrony with the host heart muscle. hESC-CMs were genome-edited to express a protein that fluoresced green with each contraction, and their behavior was studied after transplantation into an infarcted macaque heart. hESC-CMs showed 1:1 synchrony with the host heart, indicating that the graft follows the heart’s natural pacemaker, an essential result for heart regeneration.

A final question was whether this regeneration improves the function of the injured heart. To assess this, engrafted NHP hearts were studied by magnetic resonance imaging (MRI), the gold standard for assessing cardiac contractile function. As illustrated in the figure below, myocardial infarction induced a 25-point drop in left ventricular ejection fraction, the fraction of blood ejected from the heart with each beat.

Control animals receiving a saline injection showed no significant improvement at 4 or 12 weeks, as expected. In contrast, four weeks after receiving hESC-CMs, ejection fraction improved by approximately 10 points, and by 12 weeks, it had improved by a total of approximately 22 points. While the number of animals followed for 12 weeks is limited, cardiac remuscularization in this study restored ventricular function back into the normal range. In contrast, the current standard of care for myocardial infarction, including reperfusion via angioplasty, ACE inhibitors, and beta blockers, increases ejection fraction by approximately 6 points.

Restoration of Cardiac Function in NHPs by Transplantation of Human ESC-derived Cardiomyocytes (hESC-CM)

Myocardial infarction reduces ejection fraction (a measure of cardiac function), and there is no spontaneous recovery in control animals receiving a saline injection (gray). All animals receiving hESC-CM (blue) showed significant improvement by 4 weeks, and by 12 weeks after treatment, cardiac function was restored to the normal range.

In summary, structural data demonstrating extensive remuscularization of the infarcted heart in conjunction with physiologic and pharmacodynamic data provide evidence that the transplanted cardiomyocytes directly restore heart contractile function.

Engineering the cells to avoid rejection due to the host immune response to the transplanted cells

Initially, we plan to establish safety with first-in-human clinical trials of our cardiomyocyte cell therapy using immunosuppression to reduce the risk of a host immune response to allogeneic transplanted cells and the potential immune rejection. Our collaborators have studied immunosuppressive regimens in NHP by transplanting rhesus cardiomyocytes derived from stem cells into the hearts of mismatched recipient NHP. An immunosuppressive regimen was identified that keeps the allogeneic grafts alive long term and is considerably less toxic than regimens used for heart transplantation. However, an approach that obviates the need for an immunosuppression regimen has the potential to improve safety and patient eligibility. Therefore, as part of our program lifecycle we intend to switch to a hypoimmune stem cell-derived cardiomyocyte over time, as this should allow us to eliminate or reduce the immune suppression required for durable maintenance of these cells.

Addressing the risks associated with potential transient arrhythmias, or abnormal heart beats, following cell transplantation

The term engraftment arrhythmia refers to a transient period of unstable electrical activity that occurs in some species over approximately four weeks following cardiomyocyte transplantation. Engraftment arrhythmias were not observed in mice, rats, or guinea pigs (probably because their natural heart rates are too fast), but they are observed in NHPs, where they cause mild symptoms, and in farm pigs, where they cause more significant symptoms. The arrhythmias follow a stereotypical course, where they increase in frequency and duration, plateau for a variable period, and then wane until the heart has normal rhythm once again. Once the heart rhythm stabilizes, the arrhythmias seem to disappear permanently. We are exploring three ways to address engraftment arrhythmias: pharmaceutical interventions, genetic modifications to the cardiomyocytes, and adjusting the stage of differentiation of the cardiomyocytes. Electrical mapping studies in NHPs and pigs suggest that the engraftment arrhythmias originate from the site of the cell injection, likely from the injected cardiomyocytes. The stem-cell derived cardiomyocytes in these experiments are more similar to fetal cardiomyocytes than their adult counterparts, and while this favors their engraftment it potentially increases the risk for the

induction of arrhythmias. We compared the expression and function of ion channels, proteins regulating the electrical properties of cardiomyocytes, in immature versus mature cardiomyocytes. We identified multiple ion channels which showed differential activity, including HCN4, CACNA1H, and SLC8A1 which are active, as well asKCNJ2 which is inactive, in immature cardiomyocyte relative to mature ones. We subsequently assessed the impact of modulation of a variety of combinations of ion channels in animal studies. We found that PSC-derived cardiomyocytes in which the expression of HCN4, CACNA1H, and SLC8A1 has been abolished and the expression of KCNJ2 has been activated (“MEDUSA cells”) creates cardiomyocytes that lack endogenous pacemaking ability but can follow exogenously supplied electrical stimulation. MEDUSA cellsrobustly engraft but do not induce arrythmias in a pig model. We are currently testing whether MEDUSA cells can effectively restore heart function in infarcted NHPs as well as the longer-term potential impacts of modifying these ion channels in cardiomyocytes.

MEDUSA cells do not induce arrythmia after transplantation in a pig model

Wildtype stem cell-derived cardiomyocyte grafts induce arrhythmias after transplantation into pig hearts (blue circles), shown by the rise in heart rate and the increased arrhythmia burden (percentage of day spent in arrhythmia). Black circle indicates one wildtype animal that died from complications of the arrhythmia. In contrast, MEDUSA-edited cardiomyocytes (3 KO/1 OE; orange squares) do not induce arrhythmias.

Immunostaining for human cardiac muscle (human-specific slow skeletal troponin I stain, brown) demonstrates large grafts of human myocardium in the heart of a pig receiving MEDUSA cells. Thus, absence of engraftment arrhythmia is not due to absence of engraftment.

Development Plan and Key Next Steps

Our key milestones include understanding any potential impact on function and safety of these MEDUSA edits, their ability to protect longer-term from the risk of arrhythmias, completing GLP toxicology studies, and additional efficacy and safety studies in NHPs and pigs, with the goal of submitting an IND in the next several years. We plan to finalize our clinical plan based on the results of ongoing studies in NHPs using the MEDUSA edits in stem-cell derived cardiomyocytes as well as early results from humans in other settings with our hypoimmune platforms.

Our in vivo Cell Engineering Platform

Overview

In vivo cell engineering aims to treat human disease by delivering a therapeutic payload to cells inside a patient’s body to repair or control genes. Historically there have been four key challenges to in vivo cell engineering:

• to any cell (by increasing the volume of distribution),

• in a specific (for instance just T cells), and

• repeatable way (such as achieving limited immunogenicity to allow re-dosing).

Our in vivo cell engineering platform is focused on engineering fusogens that, when combined with delivery vehicles, can effectively deliver a payload to a desired cell or location in the appropriate quantities in vivo. The combination of a fusogen with a delivery vehicle referred to as a fusosome. We believe our platform provides us with the flexibility to deliver a wide range of payloads to make different modifications for different diseases, as well as delivery vehicle options to address volume of distribution and re-dosing, which could fundamentally expand the treatment potential of in vivo therapies.

Our Approach to Building our in vivo Cell Engineering Platform

We have approached the development of our in vivo cellengineering platform by investing in solutions to overcome the key challenges outlined above:

Our Approach to Building our in vivo Cell Engineering Portfolio

We have prioritized cell types for our programs where:

• high unmet need can be addressed by modifying a particular cell type;

Based on this prioritization, we are initially focused on three cell types: T Cells, Hepatocytes, Hematopoietic Stem Cells

History of in vivo Cell Engineering and Current Limitations

Starting several decades ago, the nascent field of gene therapy focused on experimenting with different means of transmitting genetic payloads via viral vectors. Seminal work by Dr. Richard Mulligan, our Executive Vice-Chairman and Head of SanaX, and colleagues established the promise of gene therapy by delivering genes into host chromosomes, thereby correcting genetic deficits. More recently, significant investments have resulted in improved safety and efficacy of viral vectors. However, most approaches continue to concentrate on adapting the innate capabilities of various viruses to transmit these payloads.

Profound benefits have been realized in cases where there is direct correlation between the biological activity transmitted by the therapy and the genetic activity that is missing in the patient. AAV vectors, which are broadly used by gene therapy researchers due to their broad tissue tropism, lack of pathogenicity, and ability to target both dividing and non-dividing cells. While these therapies have had a categorical impact on their target patient populations, they have only scratched the surface of the potential of in vivo cell engineering, with success limited to a small number of patients.

Broad impact of gene therapies has been limited by challenges within three key areas:

Payload delivery is limited by:

Genome modification is limited by:

Execution in manufacturing is limited by:

Our Solution – Fusogen Technology

To address some of the existing challenges of in vivo cell engineering, we are developing our fusogen technology by engineering proteins found in nature to enable the delivery of any payload to specific cells.

Background on Fusogens

Fusogens are a well-studied class of naturally occurring proteins that mediate the trillions of cell-to-cell and intracellular fusion events occurring in the human body every second. In 2013, the Nobel Prize in Physiology or Medicine was awarded for the elucidation of the roles of fusogens in mediating intracellular trafficking in nature. First, fusogens enable recognition of a specific target membrane. Second, they promote membrane fusion by acting as thermodynamic engines for opposing membranes, pulling them together and thereby promoting fusion.

Our Fusogen Technology

Fusogens are widely used by enveloped viruses to confer target specificity and to drive the process of introducing material in target cells. A well-known current example of a viral fusogen is the SARS-CoV-2 coronavirus that causes COVID-19. This virus uses its spike glycoprotein to target cells expressing the ACE2 receptor and to fuse with the cell membrane of host cells and release the viral genome into the cell. Many other biological processes utilizing fusogens for the delivery of complex, diverse, and large payloads to specific cell types have also been found. For example, the process of fertilization occurs as a result of a sperm fusing specifically with the egg and the transfer of the paternal genetic material to the oocyte. Similarly, the fusion of myoblasts with other myoblasts is essential for the formation, growth, and regeneration of skeletal muscle. The myoblast delivers an entire novel nucleus to the muscle cell, highlighting the utility of this system to deliver quite large and complex payloads. These and a myriad of other processes rely on this vast class of protein machines.

Applying fusogens to in vivo cell engineering

Building on both our team’s deep understanding of fusogen biology and extensive research in protein engineering, we are developing a technology designed to allow us to engineer the biological properties of these naturally occurring proteins. In doing so, we are developing a highly modular system that can specifically target numerous cell surface receptors and thereby deliver diverse therapeutic payloads to a variety of cell types.

Our current programs use fusogens derived from a virus from the paramyxoviridae family. The fusogen protein complex is comprised of two proteins: the receptor recognition G protein and membrane fusion F protein. The combination of a fusogen with a delivery vehicle such as a gene therapy vector or lipid vesicle is referred to as a fusosome. The diagram below depicts the mechanism of fusogen-mediated membrane fusion. This protein complex is found on the outer membrane of the fusosome (1). As the fusosome interacts with cells, only those with the target receptor will engage the G protein of the fusogen complex (2). The binding of the G protein to the receptor stimulates the F protein to initiate its membrane fusion activity. The F protein first partially unfolds to bind to the target membrane (3) and then refolds to bring the target and fusosome membranes in proximity (4), to ultimately promote membrane fusion (5), and subsequent payload delivery.

Mechanism of Fusogen-Mediated Membrane Fusion

The G protein has the potential to be engineered for a high degree of cell selectivity. To accomplish this, we first engineer the G protein so that its natural binding domain is no longer functional. We then add a targeting scaffold to the G protein that re-directs the fusogen to a cell-specific receptor. The targeting scaffold can be any one of naturally occurring or synthetic single chain affinity binders, such as single chain variable fragment (scFvs), camelid single-domain antibodies (VHHs), or designed ankyrin repeat proteins (DARPins). Finally, we iteratively rebuild our fusogen using insights from protein engineering to improve titers, or potency. By serially swapping different targeting scaffolds we believe we can target multiple different cell surface receptors, giving us the ability to target many different cell types.

Re-targeting the specificity of the G-protein is a challenging protein engineering problem, since altering the protein structure directly impacts all aspects of biological function. However, once we have achieved the desired specificity and potency for a certain cell type, we have the ability to deliver a variety of payloads to that cell. This feature of the technology should allow us to create multiple therapies targeting a variety of diseases with each successful fusogen. As a result, we believe success with any initial therapy targeting a given cell type could meaningfully advance lead candidate selection for other indications and increases our confidence that we will be successful with subsequent therapies targeting that same cell type. For example, a successful hepatocyte-targeting fusogen applied to a fusosome for a given monogenic liver disease meaningfully accelerates lead candidate selection and increases our confidence that we will be successful with subsequent therapies targeting hepatocytes.

Addressing key in vivo cell engineering challenges

We believe that our in vivo cell engineering platform enables us to address key challenges associated with successful in vivo cell engineering – payload delivery, genome modification, and execution in manufacturing:

Payload delivery

High cell specificity for diverse cell types. We believe we can engineer fusogens with cell specificity to maximize on-target effects, while reducing or eliminating off-target risk. In our research, we have used fusogens to successfully target numerous cell surface receptors and cell types. As an example, in preclinical studies, we have demonstrated that our fusogens can specifically target CD8, CD4, or CD3 T cells (see the subsection titled “Our in vivo Cell Engineering Pipeline—T cell Fusosome Program”), potentially enabling delivery of a payload in vivo to transduce specific T cell populations and enabling targeted cell killing through the creation of CAR T cells.

Engineering Fusogens to Target a Variety of Cell Types

Figure A depicts the increased transduction efficiency (measured in titer) of a fusogen engineered for targeting CD20 on receptor enriched B cells as compared to B cells that were negative for the receptor. Similar transduction efficiency was also observed in an engineered fusogen targeting a neuronal surface protein, GRIA4, as depicted in Figure B. Neuronal-specific transduction of the Green Fluorescent Protein (GFP) payload in the murine hippocampal region was observed using a fusosome specific for GRIA4 when injected into the hippocampal space (as depicted by the green coloring in Figure D) compared to widespread transduction when using a VSV-G fusogen (Figure C). Confirmation of neuron-specific targeting of the fusogen can be observed by the colocalization of GFP positive cells (green, Figure E) with the presence of a neuron-specific protein (NeuN in red, Figure F) and considering the high degree of overlap (colocalization seen as yellow, Figure G). Figures C-G from Anliker et al, Nature Methods, 2010.

Broad volume of distribution. Our SanaX business unit is actively working on next generation approaches to broaden the volume of distribution, including exploring cells as fusosome delivery vehicles.

Immunogenicity. We have initially focused our efforts on selecting fusogens for which the general population does not have pre-existing immunity. We are also working with a number of fusogens that exist naturally in humans, as neither these native fusogens nor re-targeted versions are likely to induce an immune response, making re-dosing more readily attainable.

Genome modification

High degree of payload flexibility. We have successfully delivered a variety of payloads including DNA, RNA, and proteins, using viral delivery methods and have used cells engineered to express specific fusogens to deliver organelles to a broad range of target cells. We believe this provides us the opportunity to potentially intervene in a wide range of human diseases.

Diverse Payload Delivery via Fusosomes

Cre protein loaded cell-based fusosomes delivered recombinase activity to cells that activated the expression of a red fluorescent protein in cells already expressing green fluorescent protein, seen as orange cells (Figures A, B). In contrast, fusosomes in which the fusogen is not included, but only contain Cre protein, showed no recombinase activity, or no orange cells (Figure C). Fusosomes loaded with fluorescently labeled RNA showed cellular localization and green fluorescence consistent with cytoplasmic delivery and translation of delivered RNA (Figures D-G). Flow cytometric analysis showed cellular uptake of fluorescent RNA (Cy5, Y axis) and GFP expression from the RNA (GFP, X axis) (Figure H). Importantly, the inclusion of a fusogen in the fusosome dramatically increased GFP expression due to the translation of the RNA. Cell-based fusosomes delivered red fluorescent mitochondria with respiration activity to cells with respiration-negative green mitochondria, (Rho0 cells) shown in Figure I. An increased oxygen consumption rate (OCR), due to respiration, was seen in Rho0 cells after Fusosome-mediated delivery of active mitochondria using two distinct fusogens (Figure J).

Expanded payload capacity. Our current fusosome has approximately twice the genetic capacity of the commonly used AAV vectors. This greater payload size increases the potential of addressing defects in larger genes or conditions where delivery of multiple genes may be required. Our research efforts include other fusosomes with even larger payload capacities. For example, utilizing a cell as the delivery vehicle can confer an almost limitless capacity.

Durability limitations. We can engineer our fusosomes to integrate into the target cell genome or to deliver non-integrating payloads. Integrated payloads allow the genetic information transmitted by the vector to be propagated durably with the genetic material of the target cell when it undergoes cell division. Thus, conditions that require this type of genetic propagation, such as genetic diseases in essential genes functioning in growing tissues or in T cells expanding after recognizing a target antigen, can be better addressed by this approach. Our preclinical studies have also demonstrated the ability to deliver gene-editing machinery, such as CRISPR, with this system. In this case, the entire payload does not integrate, but instead, it transiently delivers the machinery to

permanently modify the DNA in the target cell. Thus, we are able to make targeted, specific, and durable repairs to the genome of the target cell.

Execution in Manufacturing

Manufacturing of cell and gene therapies remains complex due to incumbent challenges in areas such as product consistency, process robustness, and scalability. Our fusosome approach has significant advantages over current solutions. Targeted delivery of complex payloads in vivo has the potential to create autologous, gene-modified cells without the complexities of ex vivo manufacturing. We believe that these therapies have the potential to have greater product consistency, improved scale, and lower costs than current autologous solutions. Currently, there are a number of therapies either approved or in development for ex vivo modification of autologous T cells and autologous HSCs. Additionally, vectors that deliver payload to random or off target cells not only create the risk for toxicities, but they necessitate meaningfully larger doses in order to ensure adequate delivery to the targeted cells. Our targeted delivery offers the potential for meaningfully lower doses, which could decrease scale needs in manufacturing.

Further, we are investing across a number of areas to improve manufacturing scale, costs, consistency, and product quality in the near-term and long-term, including by establishing and maintaining our relationships with our CDMO partners and investing in establishing and operating our own GMP manufacturing facility. Manufacturing novel fusosome compositions is complex. Since our inception, we have invested in scientific and process engineering expertise to improve manufacturing of our therapies. Examples include novel producer cell lines, novel processes and analytical technology, as well as incorporating suspension bioreactors into our process early in the research phase. By building out these capabilities early, we hope to improve the probability of technical success for our programs and have a thoughtful approach to deliver consistent supply while managing cost of goods with the goal of improving patient access.

Our in vivo Cell Engineering Pipeline

T Cell Fusosome Program (SG242, SG295, SG233, SG221, SG239)

Our most advanced CAR T cell fusosome product candidates (SG242, SG295) target CD19+ cancer cells, including NHL, CLL, and ALL. We intend to develop these product candidates with the goal of submitting an IND as early as 2022. In parallel with the CD19 CAR product candidates we are developing other CAR T cell therapies, including BCMA product candidates for the treatment of multiple myeloma (SG221, SG239), CD22 product candidates for the treatment of NHL, CLL, and ALL (SG233), as well as other targets on a spectrum of cancers.

Background on B Cell Malignancies

B cell malignancies represent a spectrum of cancers including non-Hodgkin lymphoma (NHL), chronic lymphocytic leukemia (CLL), acute lymphoblastic leukemia (ALL) and multiple myeloma (MM) and result in over 100,000 deaths per year in the United States and Europe.

NHL is the most common cancer of the lymphatic system. NHL is not a single disease, but rather a group of several closely related cancers. Over 77,000 cases of NHL are diagnosed annually in the United States and the most common subtype of NHL overall is diffuse large B cell lymphoma (DLBCL). DLBCL, if left untreated, may have survival measured in weeks or months.

CLL is the most common type of leukemia, and it occurs most frequently in older individuals, with diagnoses in people under 30 years of age occurring only rarely. Each year, approximately 20,000 patients are diagnosed with CLL in the United States. Approximately 20–25% of CLL patients initially present with poor risk disease. Median progression-free survival in these high-risk groups is often less than 12 to 18 months after frontline therapy, and less than 12 months in R/R disease.

ALL is an uncontrolled proliferation of lymphoblasts, which are immature white blood cells. The lymphoblasts, which are produced in the bone marrow, cause damage and death by inhibiting the production of normal cells. Approximately 6,000 patients are diagnosed with ALL in the United States each year, the vast majority of the approximately 1,500 deaths per year occur in adults. Approximately 80% of cases of ALL in the United States and Europe are B cell ALL, which almost always express the CD19 protein. The five-year overall survival in adults over the age of 60 with ALL is approximately 20%, and in patients with R/R ALL after two or more lines of therapy, the median disease-free survival is less than six months. B cell ALL is the most common cancer in children. Although children with ALL fare better than adults, children with R/R disease have poor outcomes. Because of the frequency, ALL remains a leading cause of death due to cancer in children.

Multiple myeloma is a cancer of the plasma cells, which typically express a protein called B Cell Maturational Antigen (BCMA). Plasma cells are B cells that have matured to specialize in the production of antibodies. Multiple myeloma is a condition in which these plasma cells become malignant, with a single clone growing at an uncontrolled pace. These myeloma cells secrete large

quantities of the same antibody, and patient symptoms can develop from the myeloma cells crowding out other plasma and bone marrow cells, leading to increased risk of infection, risk of bone destruction, and kidney disease. Multiple myeloma is the second most common hematologic malignancy making up approximately 2% of all cancers, accounting for over 32,000 new cases per year with 12,800 deaths estimated to occur in 2020 in the United States.

Current Treatment Landscape and Unmet Need

First line therapy for NHL typically consists of multi-agent cytotoxic drugs in combination with the monoclonal antibody Rituxan. In younger patients with NHL who have good organ function, high dose chemotherapy followed by stem cell transplantation is often used. Patients often relapse, however, and over the last three years, several therapeutics have been approved in the United States for the treatment of patients with R/R NHL who have received prior therapies. These approved therapies include CD19 CAR T therapies tisagenlecleucel, axicabtagene ciloleucel and lisocabtagene maraleucel, CD19 antibody drug conjugate therapy polatuzumab vedotin, and CD19 antibody tafasitamab. Recently, pivotal trials with approved CD19 CAR T cells have been shown to be superior to standard of care approaches in patients in second line R/R NHL raising the possibility of broader impact for CD19 CAR T cells for patients with NHL.

Newly diagnosed CLL patients are often treated with targeted therapies such as BTK inhibitors, PIK3 inhibitors, BCL-2 inhibitors, or monoclonal antibodies targeting CD20, or CD52 in combination with chemotherapy. However, most patients treated with these regimens become refractory. Numerous drug candidates are in clinical development for the refractory patients, including next-generation kinase inhibitors and both autologous and allogeneic CAR T therapies targeting CD20 and CD19.

Cure rates for ALL patients have continued to increase over the last four decades, with pediatric ALL cure rates reaching greater than 80% in developed countries. This progress has been enabled by advances in combination chemotherapy, monitoring of minimal residual disease, expanded use of kinase inhibitors for Philadelphia chromosome–positive ALL, and the recent approval of Kymriah for R/R pediatric ALL. Adult patients fare much worse, however, with 5-year overall survival rates of approximately 20%, and there are still significant challenges managing R/R disease across all age groups. Multiple therapeutic candidates are in development for these R/R patients, including proteasome inhibitors, antimetabolites, JAK inhibitors, monoclonal antibodies, as well as autologous and allogeneic CAR T candidates.

First-line therapy for MM is induction and high-dose chemotherapy followed by a potential stem cell transplant. There are no curative treatment options for MM patients and the standard of care for R/R MM includes immunomodulary agents, proteasome inhibitors, monoclonal antibodies, cytotoxic agents, and hematopoietic stem cell transplant. Despite the recent advancement in available therapies for disease management, the 5-year overall survival rate remains approximately 50%. To this end, several groups are investigating autologous and allogeneic CAR T cell therapies for R/R MM. BCMA is among the most promising antigens used to target MM, with two approved BCMA CAR T therapies idecabtagene vicleucel, and ciltacabtagene autoleucel, as well as multiple late-stage clinical trials ongoing. Novel treatments with other mechanisms of action are also under development, including bispecific T cell engagers, next-gen antibodies, and antibody drug conjugates.

As highlighted above, recent therapeutic advances across R/R B cell malignancies have led to a variety of treatment options and better patient outcomes. In particular, autologous surface protein directed CAR T therapies have been highly effective in certain subsets of patients with R/R disease. However, not all patients have access to novel therapies, and even with them, many patients will ultimately relapse and succumb to their cancer, resulting in 100,000 deaths per year in the United States and Europe across these indications.

There are two outstanding challenges that have limited utilization of these CAR T therapies and their impact on broader groups of patients.

Relapse. The emerging post-approval data with tisagenlecleucel and axicabtagene ciloleucel have indicated that there are two broad categories of relapse. One involves loss of CD19 on malignant cells resulting in tumor escape. This finding was initially established for ALL and is the cause of relapse after CAR T cells for roughly half of patients. More recent data indicate that low antigen expression contributes to the lack of response in a meaningful number of patients with NHL. CD19 CAR T treatments have recently been tested in pivotal trials in earlier lines of therapy for NHL, which raises the possibility of more patients being treated with CD19 CAR T therapy who subsequently relapse due to CD19 loss. Therefore, the development of CAR Ts targeting an alternate antigen beyond CD19 may be beneficial to address this growing unmet need. Data from several studies have shown that CD22 CAR T treatment has led to complete responses in NHL and ALL patients that have failed to reach a complete response or relapsed after treatment with a CD19 CAR T cell product. Additionally, dual targeting of CD19 and CD22 as the initial form of treatment may prevent this form of relapse, offering patients the potential for both a higher rate and longer duration of complete response. The second pattern of relapse relates to suboptimal CAR T cell functionality (poor expansion, poor persistence, T cell exhaustion) resulting in relapse of cancer that retains the targeted antigen. Unfortunately, re-infusion of the same CAR T cell product has had limited benefit

in these patients although treatment with a different CAR T cell has demonstrated some promise in the context of ongoing clinical trials.

Manufacturing. The manufacturing process for a patient-specific product is complex, leading to limited access due to both infrastructure and cost considerations. As such, approved CAR T cell therapies have not been available to all patients in need of these highly effective therapies. Even for patients who are fortunate enough to have access, inevitable delays (often a month or more) in manufacturing may prevent use of therapy in patients with rapidly progressing malignancies. There are groups that are seeking to overcome access limitations by using healthy donor-derived, or allogeneic, CAR T cells instead of patient T cells. This approach yields off-the-shelf therapeutics that can be manufactured consistently, but questions remain around efficacy and durability, largely due to the inability to effectively control the host versus graft response with concern for eventual rejection of these products. As will be discussed in the subsection titled “—Our ex vivo Cell Engineering Pipeline,” our ex vivo allogeneic T cell program also seeks to address this host versus graft response.

T Cell Fusosome Program

Our T cell fusosome approach provides us with an opportunity to develop potential product candidates to expand access to CAR T cell therapy to many more patients in need. In addition, we believe the ability to deliver a payload encoding a CAR to a T cell inside the body has the potential to improve effectiveness over ex vivo manufactured CAR T cell products. Experience thus far has demonstrated that both CD8+ and CD4+ T cells contribute to the CAR T cell response. Thus, the fusosome programs we are developing will deliver the CAR gene using fusogens that directly and specifically target the CD8 co-receptor or the CD4 co-receptor on T cells following a single intravenous injection. These approaches could result in the generation of therapeutically active CAR T cells without the complexities and delays associated with the process of T cell collection and ex vivo manufacturing. Furthermore, the ex vivo expansion in the presence of high cytokine concentrations, while necessary for the manufacture of approved CAR T cell products, also contributes to marked changes in T cell quality that may not be therapeutically beneficial. We believe the generation of a CAR T cell within the natural physiological environment has the potential to improve the quality of the CAR T cell generated, potentially improving both efficacy and the side effect profile. Finally, the effectiveness of ex vivo manufactured CAR T cells is dependent on the administration of a lymphodepleting preparative regimen prior to infusion to facilitate expansion of the CAR T cell product, which can have meaningful adverse safety implications. We do not expect to need a lymphodepleting regimen prior to in vivo delivery of the CAR gene, as our goal is to expose our fusosomes to as many T cells in the body as possible.

Preclinical Data

Our preclinical data have demonstrated that fusosomes can deliver a genetic payload specifically and efficiently to human T cells in culture and in immunodeficient mice with intraperitoneally injected human peripheral blood mononuclear cells (PBMC) and fused with a single dose of a fusosome. The T cells can be categorized into functional subsets based on the expression pattern of cell surface molecules. CD3 is a protein expressed on all T cells, CD4 is expressed on the Helper T cells that primarily activate T and B cells to carry out their function, and CD8 is found on cytotoxic T cells that primarily kill cancerous or virally infected cells. We generated fusogens against these three cell-surface molecules and have demonstrated that we can deliver a marker gene to cells bearing these cell surface proteins in vitro.

Fusogens Demonstrate the Ability to Target Multiple T Cell Subtypes

Fusosomes containing a gene that encodes a fluorescent marker protein called GFP (used to identify cells have been genetically modified by the fusogen) can efficiently and specifically deliver GFP to T cells in culture (CD8, CD4, and CD3). Expression of GFP is restricted to the population of T cells that express the specific T cell receptor targeted by the fusogen (CD8, CD4, or CD3).

We have further established that fusosome delivery of a CD19 CAR gene to CD4 or CD8 T cells results in killing of human B cells and CD19+ leukemia cells in culture:

Delivery of CD19 CAR to CD4 T Cells Leads to in vitro Killing of B Cells and CD19+ Leukemia Cells

Demonstrates that the fusosome-generated CD4 CAR is functional and eradicates both nonmalignant B cells (CD19+/RFP-) as well as CD19+ leukemia cells expressing NALM6-RFP.

We have also validated, in vivo, the tumor-killing activity of CD8 T cells to which CD19 CAR has been delivered via a fusosome.

Delivery of CD19 CAR to CD8 Cells Leads to in vivo Killing of Leukemia Cells in a Human Xenograft Mouse Model

Left panel: demonstrates activity of CD8 fusosome delivering CD19 CAR to human T cells in a murine leukemia xenograft model (Nalm-6). Note that when compared to untreated controls, fusosome delivery results in eradication of leukemia cells. Activated T cells were cultured with CD3/CD28 beads for 3 days prior to injection. CD8 fusosome delivering the CD19 CAR is effective regardless of activation status of T cells at time of injection. Right panel: represents quantification of luminescence (representing leukemic burden) from mice shown in left panel. Both cohorts of fusosome treated mice had significantly reduced tumor burden when compared to control as early as D10 (p≤.0001; One-way ANOVA Bonnferroni) Experimental note: Tumors injected on Day Zero, Donor T cells injected on Day Three and Fusosome injected on Day Four.

Using a human xenograft mouse model for leukemia (Nalm-6), we observed both prolonged survival and clearance of the leukemic cells. During the manufacture of autologous CAR Ts, cytokine signaling has to be activated in order to successfully produce functional CAR T cells. In our mouse experiments the CD8 fusosome was able to generate CD19 CAR cells just as effectively with activated as non-activated donor T cells.

Several of our human T cell fusogens cross-react on non-human primate (NHP) T cells including our lead candidate CD8 fusogen. We have used the fusogen to deliver a CD20 CAR into six NHPs (the CD20 CAR was chosen as the CD19 CAR to be used for our clinical programs does not cross-react with NHP B cells). As shown below, a single intravenous administration of our CD8 fusogen containing a CD20 CAR was associated with B cell depletion, including in the blood and in lymph nodes, in four out of six NHPs, occurring between day seven and fourteen. This result is consistent with reported observations that peak expansion of ex vivo manufactured CAR T cells typically occurs during the second week. Serum cytokines were transiently elevated at day seven in all fusosome-treated animals. CD20 CAR transgene (by vector copy number) and CD20 CAR mRNA could be detected in peripheral blood between days three to ten, and in the spleen at study termination. Importantly, there was no infusion-related toxicity or evidence for CAR-associated toxicity (cytokine release syndrome or neurotoxicity) other than the intended B cell depletion. The ability to deliver fusogen without toxicity and with evidence for activity in NHP are critical milestones for the program. In addition, the NHPs received no T cell activating agent or lymphodepletion. The latter potentially supports a path for the fusogen platform to enable delivery of CAR therapy without the lymphodepletion regimens used by existing ex vivo approaches, which have toxic side-effects. This and future NHP experiments will also provide important information on dosing parameters, durability of the effect, and provide pharmacokinetic, pharmacodynamic, and toxicology data.

Delivery of CD20 CAR to CD8 Cells Causes B Cell Depletion in NHPs

Graph demonstrates activity of CD8 fusosome delivering CD20 CAR in NHPs. CD8 fusosome was delivered as a single intravenous infusion. B cell counts were quantified in peripheral blood using flow cytometry for CD20+ cells. Given expected NHP-to-NHP variability in baseline counts, data is represented as single lines per NHP and is shown as deviation from baseline level for each NHP at Day Fourteen (when maximal expansion of CAR T cells are expected). Significant B cell depletion is observed in four out of six NHPs. Note that no T cell activation was provided prior to fusosome delivery.

Development Plan and Key Next Steps

We are currently conducting experiments to validate the ability of a systemically administered fusosome to transduce T cells in an NHP and for these CAR T cells to deplete B cells. These NHP studies are also expected to inform preclinical pharmacology and toxicology.

As a next step, we intend to conduct our in-life GLP toxicology studies. While these studies are ongoing, we intend to scale our GMP manufacturing and finalize our initial development plan. We intend to submit an IND for SG295 to support clinical trials in patients with NHL in 2022. We continue to advance additional programs and plan to submit INDs over the next several years: SG242 in NHL, SG233 which delivers a CAR gene targeting CD22 and could be combined with targeting CD19, and our BCMA programs in MM.

Hepatocyte Fusosome Program

Numerous genetic metabolic diseases arise from gene defects that manifest in the liver and, in particular, in the hepatocyte. Additionally, hepatocytes can serve as protein manufacturing sites to deliver proteins to other cells in the body. Multiple modalities exist that enable delivery of genetic material to liver cells, including AAV and LNPs. However, these approaches have limitations, including non-integrating payloads, payload size, lack of cell specificity, and, in the case of AAV, immunogenicity. Our fusogen technology, which we expect will be able to deliver a payload specifically to hepatocytes in the liver, has the potential to address these limitations. Success with this hepatocyte-targeting technology may allow us to generate therapies for a number of genetic disorders. We are developing our lead product candidate, SG328, for ornithine transcarbamylase (OTC) deficiency, and we expect to submit an IND in the next several years.

Hepatocyte Targeting Capability

Targeting the hepatocyte with a fusogen can enable specific delivery of either integrating or non-integrating payloads. It can also be used to deliver the machinery of gene editing and gene modification tools to these cells, both with or without the inclusion of DNA to replace a mutated gene or gene fragment. Since we anticipate that hepatocytes transduced with fusosomes will harbor the novel genetic construct in their genome, all progeny of that cell will also have the genetic construct. Thus, the natural turnover and organ growth will not dilute the genetic construct, providing the potential for long-term expression and efficacy even when the

fusosome is delivered during infancy, childhood, or when it is delivered to treat a disease where the disorder can cause rapid hepatocyte turnover.

We believe that success with an initial hepatocyte-targeted fusosome will meaningfully accelerate our future hepatocyte programs. Once a hepatocyte-targeting fusosome is established, our subsequent programs will require only substituting the relevant payload to correct for the defective gene in question, opening up the possibility to address multiple inherited liver diseases. Our initial focus is on monogenic diseases with clear biology linking the missing activity of a gene in hepatocytes to a disease outcome. According to the National Institute of Health, over 30 genetic disorders of the liver exist, impacting over 10,000 births annually around the world. Many of these disorders lead to death in the first few years or cause long term disabilities. Proof of concept within this set of initial diseases will enable expansion to other diseases, such as hemophilia, where we may be able to address an unmet need by providing a durable in vivo therapy in the hepatocyte.

Preclinical Data

Our ability to use our hepatocyte-targeting fusosomes in relevant animal models is limited by a lack of cross-species transduction. To address this, we first developed murine disease models and introduced the therapeutic payload utilizing a conventional lentivirus pseudotyped with VSV-G. The VSV-G fusogen targets the LDL receptor which is highly expressed in hepatocytes providing a potent in vivo delivery vehicle for hepatocytes. These models established proof of concept to treat these diseases through genomic integration of the corrected gene as well as transduction efficiency in the range that would be needed for efficacy. However, the LDL receptor is found on a significant number of other cell types resulting in extensive off-target transduction. We do not intend to move forward with this non-specific lentivirus construct, as its lack of specificity creates potential challenges in humans. However, it does provide a preclinical model system for us to understand the percentage of hepatocytes, as well as the expression level of the novel genetic material that are required for the intended therapeutic effect.

In parallel, we have developed and improved hepatocyte-specific fusosomes for high on target transduction efficiency (as measured by titer), with the goal of achieving potency comparable to or better than what we see with conventional lentivirus. Engineering of hepatocyte specificity is generated through the choice of target receptor selectively expressed in human hepatocytes. Through an iterative process focusing on multiple hepatocyte-selective cell surface protein targets, diverse binders, and protein engineering, we have developed constructs that have met our potency goals. We have tested these constructs in vivo in mouse models and shown that they can transduce human hepatocytes at levels comparable to conventional lentivirus with significantly lower frequency of off-target transduction. Furthermore, the transduction occurs in a dose-dependent manner.

Improved Targeted Fusogens Exhibit High in vitro Transduction of Primary Human Hepatocytes (PHHs)

Improved hepatocyte-targeted fusosomes show levels of in vitro transduction similar to conventional lentivirus. Protein engineering of Fusogen 1b resulted in new sets of fusogens with significantly increased titer on PHHs. The most potent of these fusosomes approach the hepatocyte titer of conventional lentivirus.

Improved Targeted Fusogens Can Transduce Hepatocytes in vivo in a Humanized Liver Mouse (FRG) at Comparable Levels to Conventional Lentivirus and in a Dose-Dependent Fashion

Hepatocyte-targeted fusosomes show levels of in vivo transduction similar to conventional lentivirus and dose dependence. Fusosomes were delivered by intravenous injection (tail vein) into humanized liver mice (FRG, or FAH- immunodeficient mice, repopulated with primary human hepatocytes). Dose for lentivirus was 1.4 1011 TU/kg (# of transducing units per kilogram mouse body weight, tested on primary human hepatocytes). Highest dose for the hepatocyte-targeted fusosome was 1.1 1010 TU/kg (1X) and lower doses were at one-third (0.33X) and one-tenth (0.1X) of the highest dose. Liver cells were harvested from injected mice seven days after injection, separated into hepatocytes and non-parenchymal cells (non-hepatocytes) and analyzed for GFP expression and species (human or mouse).

Background on OTC Deficiency

OTC deficiency is the most common inherited disorder of the urea cycle, the process by which the body detoxifies ammonia and produces urea. It is the only urea cycle disorder that is X-linked, leading to more severe disease in males. OTC deficiency occurs in approximately 1 in 50,000 births, and there are approximately 10,000 patients worldwide. A deficiency of the OTC enzyme leads to accumulation of ammonia, which can lead to neurotoxicity manifesting early as vomiting and anorexia, before progressing to a progressive lethargy, seizures, intellectual impairment, coma, and death. The severity and age of onset of OTC deficiency can vary with the most severely affected, typically males, presenting shortly after birth. In this severe, neonatal onset of OTC deficiency, patients present with an overwhelming illness that rapidly progresses with up to 90% mortality rate despite advances in standard of care treatments. In less severely affected patients who present later in childhood or as adults, severe elevations of ammonia and resulting neurotoxicity still occur, primarily precipitated by an illness or excessive protein intake.

OTC Deficiency: Current Treatment Landscape and Unmet Need

The standard of care for patients with OTC deficiency includes a low protein diet, nutrient supplementation, and the use of ammonia scavengers such as benzoate, phenylacetate, or phenylbutyrate. Despite all of these measures, patients may still experience acute hyperammonemia crises particularly in the setting of increased protein catabolism that can be induced by viral illness or certain medications. These acute crises are treated with supportive care including kidney dialysis for rapid ammonia reduction. The frequency and duration of hyperammonemia crises has been directly linked to poor long-term outcomes and intellectual disability. The only curative therapy available is liver transplantation, which has become more common as surgical techniques and supportive care have improved over time. In those patients with severe, neonatal onset of OTC deficiency, liver transplantation is commonly performed before the age of five and, in some cases, can occur before one year of age.

In addition to the standard of care therapies noted above, therapies to replace the defective OTC gene have been pursued. Recent trials have primarily utilized AAVs to deliver a corrected OTC gene. While these viruses have to date been generally well tolerated, they are still associated with significant immunogenicity that can preclude use in the up to one third of patients with pre-existing antibodies to AAV and can lead to systemic symptoms, including elevated liver enzymes. Beyond the challenge of pre-existing antibodies, the primary drawback is the potential for transient efficacy as the gene replacement via AAV would not be expected to be

permanent if replication of the target cell occurs. While the durability of an AAV delivered gene replacement for OTC deficiency depends on many factors, one of the key determinants is the rate of hepatocyte turnover. This is especially relevant in pediatric patients with growing livers and rapid cell turnover. This dilution of effect has been supported by animal studies where AAV delivered gene replacement was successful in adult animals but not successful in younger animals. The most severe form of OTC deficiency presents in the neonatal period and, if a donor is available, may be treated with liver transplantation, a permanent gene therapy that provides long lasting benefit to patients is required to address the greatest unmet need from OTC deficiency. Additionally, as hepatocytes continue to divide approximately once a year even in adults, a durable gene therapy would also be expected to provide an advantage even in adult patients where an AAV delivered gene therapy is likely to lose function over time.

We believe our approach of pursuing a permanent gene replacement therapy has the potential to improve morbidity, mortality, and quality of life even in the youngest, most severely ill patients.

Development Plan and Key Next Steps

We are conducting mouse studies to establish proof of concept and inform the dose profile of our lead hepatocyte fusosome. In the near term, we are seeking to finalize the hepatocyte-targeted fusosome candidate and begin GLP production. Dose and safety of our lead fusosome compositions for OTC will be further informed through NHP studies, and with success, we expect to submit an IND in the next several years. The hepatocytes of the liver are the one cell type readily accessible to competing technologies such as AAVs and LNPs. Each of these technologies has limitations overcome by the fusosome platform, but any future investment in therapeutics targeting hepatocytes will need to consider if these fusosome advantages are meaningful enough to patients to justify the investment. For example, we are currently evaluating whether our technology offers differentiated solutions to these liver-targeted gene therapies for genetic diseases such as Hemophilia and Alpha-1 antitrypsin deficiency.

HSC Fusosome Program

We are developing hematopoietic stem cell (HSC) targeted fusosomes, designed to target and repair genetic abnormalities underlying diseases such as sickle cell disease and beta-thalassemia (SG418), with the goal of achieving preclinical proof of concept as early as 2023.

Background on hemoglobinopathies

Devastating inherited hematologic disorders, including sickle cell disease, beta-thalassemia, and other hemoglobinopathies, are caused by a monogenic variant, and patients suffering from these diseases are candidates for in vivo cell engineering.

SANA US Equity

SANA · 10-K · period ended 2021-12-31

Item 1. Business.

SANA US Equity☆

SANA · 10-K · period ended 2021-12-31

Item 1. Business.

SANA US Equity