Morgan Stanley

Morgan Stanley February 29, 2016 Editas Medicine Starting A Gene Editing Revolution Industry View Stock Rating In-Line Equal-weight Price Target $28.00 We initiate at EW with a $28 PT; Despite our LT optimism for the promise of gene editing, two near-term headwinds are likely to keep EDIT range bound - IP uncertainty and a lack of clinical catalysts until 2017. Over the long term, however, we see Editas as the premier gene editing company. We initiate at EW with a $28 PT: Editas is using a new technology called CRISPR/Cas9 to edit (replace/remove) bad genes with good genes and achieve a therapeutic benefit in patients with genetic diseases. This approach differs from the more commonly known gene therapy in that it can accurately remove and replace genes instead of inserting a new gene sequence without specificity as happens in gene therapy. Thus, the potential of CRISPR is broad with -6,000 genetic diseases of which less than 5% have treatments. At odds with our positive long-term view of Editas and CRISPR technology are two headwinds which we see as keeping EDIT in check: (1) Over the course of the next 1-2 years Editas and its academic parters are going to engage in a substantive IP battle over CRISPR technology versus other companies, namely, Intellia and Crispr Therapeutics, which is likely to limit stock appreciation; and (2) initial clinical data is unlikely to be available until late 2017 at the earliest, limiting any significant derisking events for the platform. CRISPR is a compelling technology and we see Editas as best positioned to realize its potential: We have high hopes that the potential for CRISPR can be translated into clinical benefit across a wide variety of diseases. Key to our optimism is that CRISPR has been used in many labs across many academic institutions with strickingly similar outcomes, suggesting that many of the key technical hurdles in small systems are understood. We believe Editas has a strong analytic process dedicated to addressing each of the main components of development. Thus, we ultimately see Editas as being successful in delivering therapeutic candidates. Importantly, for the lead clinical program in Leber's disease, we see it as the right first target given that it is relatively easier to deliver therapy to the photoreceptors, and the closed system of the eye limits potential safety risks. IP a key area of debate over the next 1-2 years: Editas, along with its academic partners from whom it licensed its foundational IP, is currently engaged in a interference proceeding to determine which claims, if any, from its IP can stand versus the competing party (University of California and its licensees). This proceeding is likely to play out over the course of the next 1-2 years. While we expect the outcome is likely to be one which requires both sides to share IP, there are tail scenarios which could lead Editas to have no freedom to operate. Editas Medicine] February 29. 2016 MORGAN STANLEY RESEARCH MORGAN STANLEY & CO. LLC Matthew Harrison '1212 761-8055 David N Lebowitz, MPH. CFA '1212 761-0324 Cyrus Amoozegar, M.D., Ph.D. .1212 761-6009 Editas Medicine ( EDIT.O, EDIT US ) Biotechnology / United States of America Stock Rating Industry View Equal-weight In-Line Price target $28.00 Shr price, close (Feb 26, 2016) 527.49 Mkt cap, curr (mm) 5133 52-Week Range $2940.12,57 Fiscal Year Ending 12/14 12/15e 12/16e 12/17e ModeiWare EPS (S) (4.79) (233) (1.16) (L81) Prior ModelWare EPS (1) P/E NM NM NM NM Consensus EPS ($)1 Div yld (%) Unkst oihemiie noted, all metrics am based on Mogan sanity mocloWaie frarnenork - ContanT.os data ,s pronle:l by Thomson Rauuns E clm ems a - Montan Staab, R{,{.I0.4 mimamt QUARTERLY MODELWARE EPS (S) Quarter 2015e 2014 Prior 2015e 2016e Current Prior 2016e Current Q1 (1.00) (0.70)a (0.21) Q2 (1.00) (0.68)a (0.24) Q3 (1.36) (0.69)a (0.29) Q4 (1.89) (0.26) (0.41) • Macron Sun by Research estmaiet a • ActualCompany /opened data Morgan Stanley does and seeks to do business with companies covered in Morgan Stanley Research. As a result investors should be aware that the firm may have a conflict of interest that could affect the objectivity of Morgan Stanley Research. Investors should consider Morgan Stanley Research as only a single factor in making their investment decision. For analyst certification and other important disclosures, refer to the Disclosure Section, located at the end of this report. EFTA01100261 Morgan Stanley Editas Medicine] February 29. 2016 MORGAN STANLEY RESEARCH Lack of near-term clinical data and IP dispute likely to keep EDIT range bound: Despite our positive view of CRISPR, we see the lack of near-term derisking clinical data as limiting potential upside to EDIT in the near-term. Further, while we expect the outcome of the IP dispute to be some sort of cross licensing agreement similar to what we have seen with antibodies, the tail scenarios present sizable risk and thus are likely to also limit upside. 2 EFTA01100262 Morgan Stanley Editas Medicine I February 29. 2016 MORGAN STANLEY RESEARCH Risk Reward Demonstration of safe administration in humans, IP resolution, and LCA10 success drive risk/reward BO 70 00 $7200(.210%1 50 40 30 $n t3 20 10 0500(. 0 Feb-14 144.14 F4045 Aug.15 Feb-115 /04.03 FS-17 Ter Psi') Source: Thomson Reuters. Morgan Stanley Research Price Target $28 Bull DCF Base DCF • Oen Ste Pis Our PT is derived from a DCF that uses a 15% discount rate and a 0% terminal growth rate beyond 2032E. $72 Editas is able to develop and commercialize therapies that receive widespread uptake. The LCA10 therapy launches in the US and EU in 2024E, treats and cures —80-90% of the addressable LCA10 population, commands premium pricing and generates peak sales of -4130M in both the US (2026E) and EU (2027E). Editas is also able to realize -568, in sales by 2032E from additional therapies in CAR-T applications, non-malignant hematology, DMD, and CF. $28 The LCA10 therapy proves successful, but additional therapies obtain modest success. The LCA10 therapy launches in 2024E in the US and EU. and treats and cures —70-80% of the addressable LCA10 population to generate -585M in peak sales in both the US and EU (2028E). While Editas is able to develop successful therapies for CAR-T, hematology, DMD, and CF, it achieves lower market share for total additional annual re✓enues of -438 by 2032E. Bear $5 DCF (Cash/share) Pipeline programs fail. Editas is not able to commercialize any therapies. and the resulting valuation is cash / share. Investment Thesis ■ We are Equal-weight Editas Medicine. The CRISPR gene editing platform has been derisked from an operational standpoint, and can be geared towards numerous disease targets. Howe✓er, the delivery and long-term safety of administering dinically relevant doses in humans needs to be proven. ■ Editas has a systemic, modular approach that may allow for differentiation over competitors ■ The first disease being targeted (LCA10, an inherited retinal dystrophy) lowers the initial risk of proving that gene editing can work in humans, as the eye is immune privileged and provides a small, contained area in which sufficient quantities of vector can be delivered safely ■ Overall, we see significant long-term potential for Editas, but remain equal-weight while initial therapies are derisked in the dinic and the IP battle plays out over the next 1-2 years. Key Value Drivers ■ Resolution to IP interference proceedings ■ Progressing the LCA10 program into the dinic ■ Advancing current discovery stage programs into the clinic such as for hematologic diseases and genetic diseases of the lung & liver Potential Catalysts ■ Initiation of Phi LCA10 trial in 2017 ■ Interference proceedings in 2016/2017 ■ Entering additional therapies into the clink in 2017/2018 Risks to Achieving Price Target ■ IP outcome that limits Editas' freedom to operate ■ Development risk associated with early nature of pipeline, and the timeline is long to initial data (first data in humans is expected from the LCA10 Phi study in 2017) ■ Competitors that could influence investor perception of the stock 3 EFTA01100263 Morgan Stanley Editas Medicine I February 29. 2016 MORGAN STANLEY RESEARCH Investment Case Summary & Conclusions We are initiating coverage of Editas with a 528 PT and an Equal-weight rating. Our rating is based on two near-term factors at odds with our longer term view of CRISPR as a platform: (1) Over the course of the next 1-2 years Editas and its academic parters are going to engage in a substantive IP battle over CRISPR technology versus other companies, namely, Intellia and Crispr Therapeutics which is likely to limit stock appreciation; and (2) initial clinical data is unlikely to be available until late 2017 at the earliest, limiting any significant derisking events for the platform. Thus, while we continue to view CRISPR as one of the more compelling next-generation technologies to address a wide variety of disease targets, we think the stock is likely to remain range bound ahead of clarity on both IP and clinical activity. Note: The authors of this material are not acting in the capacity of attorneys, nor do they hold themselves out as such. This material is not intended as either a legal opinion or legal advice. The information provided herein does not provide all possible outcomes or the probabilities of any outcomes. The result of any legal dispute or controversy is dependent on a variety of factors, including but not limited to, the parties' historical relationship, lows pertaining to the case, relative litigation talent trial location, jury composition, and judge composition. Investors should contact their legal advisor about any issue of law relating to the subject matter of this material. Key Investment Points 1. CRISPR is a compelling technology and we believe Editas is best suited to translate the technology into therapeutic benefit - We are positive about the potential for CRISPR to be translated into clinical benefit across a wide variety of diseases. Key to our optimism is that CRISPR has been used in many labs across many academic institutions with similar outcomes, suggesting that many of the key technical hurdles in small systems are understood. That said, there are still many key risks and challenges that need to be overcome including identifying efficient edits for each target, appropriate delivery to the tissue, expanding the platform's applicability across a range of different types and kinds of gene edits and various other engineering challenges. However, we believe Editas has a strong analytic process dedicated to addressing each of the main components of development. Thus, we ultimately see Editas as being successful in delivering therapeutic candidates. 2. IP battle will remain an overhang, but we continue to see cross licensing as the most probable outcome - The P in the CRISPR space is complicated, varied and nuanced. Thus, it is not surprising that various academic institutions believe they each have foundational P. Early in January 2016, foundational patents which Editas has licensed from Harvard, the Broad Institute and MIT were named in a patent interference proceeding with the University of California, University of Vienna, and Emanuel Charpentier. These patents are held by Caribou Biosciences and licensed to Intellia. Crispr Therapeutics also has rights to the same IP as Intellia though the second scientific founder, Dr. Doudna. The debate between the parties, which we discuss further in this report, is whether the initial discovery of CRISPR in prokaryotic cells is easily translated into mammalian cells (Broad was first to discover in the latter category while UoC in the former). Given the complexity of the P — 11 patents are named in the interference —we believe a plausible outcome is one where certain claims in Editas IP are narrowed (right now Editas is the only company with granted IP) and certain claims in the Caribou P are granted, but overall both parties would need rights to the other party, similar to the situation which developed from the foundational antibody IP. We assume modest royalties which generally cancel each other out. Nonetheless, the uncertainty created by this situation — including the fact that one outcome could be that Editas would have no freedom to operate if all its IP was overturned — is likely to keep EDIT range bound as the interference proceeds. 4 EFTA01100264 Morgan Stanley Editas Medicine] February 29. 2016 MORGAN STANLEY RESEARCH Exhibit 1: Indicative Timeline for IP Interference Proceedings 12 months -C14 days 4-6 weeks 6 weeks 3 weeks 6 weeks 6 weeks 3 months 3 months Declaration of Interference Initial Papers Motions Lists I Call with APJ TP1 Motions TP2 Responsive Motions TP3 Oppositions TP4 Replies Oral Argument Decisions on Motions Cross-Examination Cross-Examination Period for Observations Motion to Exclude, Records Source. Company Dili Morgan Stank] ROttilth 3. Clinical data is the key upside driver, but is not due until late 2017 at the earliest- Editas and its CRISPR platform has the ability to target a wide variety of diseases with over 6,000 diseases caused by a genetic mutation and 95% having no approved therapy. Key targets include muscular dystrophy, cystic fibrosis, various malignant and non-malignant hematologic diseases and other liver direct targets. Thus, the universe of available diseases is large; however, all of these programs are currently in preclinical development and unlikely to move into the clinic in the near-term. The most advanced program is for a eye disease called Leber congenital amaurosis (LCA), which impacts a patient's retina and leads to blindness. This program is likely to enter clinical testing in 2017. Importantly, we view this as an appropriate target for an initial clinical program. Based on the literature -10-20% of the photoreceptors in the eye need to be edited to produce a therapeutic response (i.e., maintenance of some vision) and since the eye is a closed body, both delivery is likely achievable and off target toxicity is likely to be limited. Thus, we see the overall risk to be more modest with the LCA10 program. Despite our positive view of this program, the timeline is likely to limit stock appreciation given clinical data is unlikely until late 2017. 4.Business development has the potential to be a near-term upside driver - Editas has already completed one licensing deal with Juno in the CAR-T space. While we think that partnership can drive value for both companies, the timelines are not near-term. Further, we have seen other companies like Crispr Therapeutics strike partnership deals with Vertex and Bayer. Thus, there is clearly interest in CRISPR therapies by larger companies. Unlike Crispr Therapeutics, we believe Editas' management strategy is better suited to preserving shareholder value and would expect management to target deals where the assets are well defined (i.e., no open-ended target deals) and either the time to market or the basic knowledge in the therapeutic category (i.e., neurology) could be accelerated by the larger party. Thus, we do expect business development and expect it to be positive for EDIT, but the timing is hard to predict 5 EFTA01100265 Morgan Stanley Editas Medicine I February 29,2016 MORGAN STANLEY RESEARCH Exhibit 2: Correction of mRNA Expression in Cells from LCA10 Patients 0 • Normal MMutant Conti'01 Source:Company DaIs Guide Pair 1 Guide PaIr2 Exhibit 4: Recent Gene Editing Deals bed say- Valet Arno Introor Exhibit 3: Editas Pipeline Program Target Gene Star Lebo Congenial Amurtosis tO CEP290 Dotorery. SC over". Nude, a 2015. nil „I . 2017 Olitett ilt4 Meagre CuMiteinter M Eye Discern Geoe Edo% lareelleto Weal OMNI Maple Deccreary NoorMairant Hereleloge 04•••••n M.Itple DriCevery Gent% etteeeesolMescle MJIple Decoverf Gera& Damsel De Lung Maple Discovery aeons innesousnmesoi Ole MS Maple Drecourf Source: Company Data. Morgan Stanley Research f tertine Pseinn C°Mbar awn Occur Annourue ,ntnt per 1),,bon of CellAberstot upfront rot .is mfortomr Royally murbortrer Allot therontaA‘ IV to address Wood *tureen. lAndens. and rceutoolal neon deem 1///1/201S foremmestt 13001Ar ton/ ern lie yea. OM MA NA DI SA, mums SlOOPA• At eeruhr creceON owned SO/SO by Saw and CeRlet thempean us ant, Ateurret 5394 neer o (Mee Moseys Cute. thoyertuuct error focus on cystic homy, and wire tell Away c...blorytann additional (hunts torpsnon lour wars 510501 54204A Vt. 111 SIOSM uelrotmometts el Sore Ea* end SAW merry ID Venn venture KO% et claelownom rout to trestmorts olormied bIrVer" tertat CAR 1 art tat cancer Morrom S/17/1015 (never. SRCOSA rn heed runnel on 522M in reiteer trIPPOn Crter Are yrA coill•boraion teethe tale1 murk. tncireenng CAAts ore nemAtoporesrt %tern ten 1/7/7015 Orreymrs Not Onflosed Nei dechsted Yes III MD hordes II)tiovarm atones onprol cowry slake in Welke Source Company Data. Morgan Stanley ROttalCh S.Given the early stage of development Editas is tough to value, but we think a unique approach is warranted - We value Editas using a two pronged approach. As is typical, we value the defined clinical programs - in this case LCA10 - and assign a value. However, separately, we also try to value the potential upside associated with the platform. Here our approach is unique as we take our revenue models for DMD, Cystic Fibrosis, CAR-T and non-malignant hematology and assume Editas could penetrate that market starting in 2025 and reach peak share of 30% of that market by 2030. We than assume a 10% probability of success as a way to gauge the potential of the platform across a wide variety of targets and diseases. 6 EFTA01100266 Morgan Stanley Editas Medicine I February 29. 2016 MORGAN STANLEY RESEARCH Exhibit 5: Bear to Bull Case Bridge for Editas Valuation Oral Sou r<* Ngorgan Stanley Reseagch Key Upcoming Catalysts 2$ Hale arse A.:39 LCAIG 72 Ilu The key upcoming catalysts for Editas include updates to the interference proceeding, updated preclinical data at the ASGCT conference in May and an IND filing on its first clinical candidate for LCA10. Obviously, general developments in the CRISPR space are likely to occur as well and could have an impact on Editas. Exhibit 6: Editas Catalyst Calendar Milestones Timing Updated preclinical data at ASGCT 2016 May 4-7, 2016 Potential new business development 2016 Potential interference proceedings and updates 2016/2017 Initiation of Phase I LCA10 study 2017 Additional therapies enter the clinic 2017/2018 End of Juno collaboration 2020 ou roe: Company Data. Morgan mangey lirnatth 7 EFTA01100267 Morgan Stanley Editas Medicine I February 29, 2016 MORGAN STANLEY RESEARCH Valuation Exhibit 7: DCF drives valuation ac Oct NO, 0.0 ae 001, On0 a.m 'on on nu ion inn ion isn a an out aon aro ant an (% on ma on n. M, to pn Sill II" Source: Company Data, Morgan Stanley Research Sum of Discounted CF ($3.4) $1,009 Net Cash $203 Equity Value $1,213 Equity Value Per Share $28 Discount Rate 15% Terminal Growth Rate 0% Time of Valuation 2016 Shares Outstanding (millions) 43 Our S28 price target includes -43.38 in peak (2032E) global revenue. We derive our price target from a discounted cash flows (DCF) analysis that uses a WACC of 15.0% and a terminal growth rate of 0%. Valuation Methodology: We prefer the use of a DCF analysis to value biotechnology companies. Given the defined patent life for each product, we believe a DCF fully captures both the upfront investment period as well as the long-term earnings power. While investors do look at biotechnology on a multiples basis (PIE), we prefer a DCF as it is more rigorous and requires more explicit assumptions about the long-term prospects of a company. Discount Rate: We use a 10% discount rate for all commercial companies, a 12.5% rate for companies with randomized Phil data and a 15% rate for all development stage companies. Given the stage of development we use a 15.0% discount rate for EDIT. Terminal Growth Rate: We model explicit revenues through 2032E with a 0% terminal growth rate Revenue: We model 5-25% penetration of the LCA10 market in the US and 3-17% in the EU starting in 2024. We assume $1M pricing in the US, increasing 1.5% annually. We assume a $900k price in the EU, increasing 1% annually. For other products we assume a blend of revenues from CAR-T, non-malignant hematology, DMD, and CF with the company capturing a small fraction of sales and having a 10% probability of success. Total estimated peak sales (2032) are -$3.38. Economics: Editas maintains worldwide commercial rights to many of its product candidates, except for their CAR-T program. We expect the company to retain global rights and launch their products themselves. COGS: We model COGS as a continuous 17%. Operating Expenses: We assume R&D of $25M in 2016E growing to -$80M in 2021E and -5250M in 2032E. R&D growth is expected to be aggressive in the next few years to account for the simultaneous development of several products. We assume SG&A of -$25M in 2016E, growing to -556M in 2021E and -$210M by 2032E, assuming product launches starting in 2024E. Key Risks To Our Price Target Include: (1) Lack of freedom to operate driven by losses int he interference proceeding around CRIS PR IP; (2) Inability to deliver CRSIPR candidates to the correct tissue or inability to achieve high editing efficiency; (3) Lack of efficacy with initial clinical data or unknown safety. 8 EFTA01100268 Morgan Stanley Editas Medicine] February 29. 2016 MORGAN STANLEY RESEARCH Debate 1- What is CRISPR/Cas9? Overview: CRISPR is a new technology that may not be familiar to many investors. While the new wave of genetherapy companies - where a gene sequence is inserted into a viral carrier (typically lenti or AAV) and then incorporated into the body's genes - has educated investors on that technique, many have heard about CRISPR, but do not know all of the details. CRISPR is a protein-RNA complex where guide RNAs locate a specific gene sequence and then allow cas9 nucleases to "edit" that sequence. It is a precise way to edit a specific gene sequence, potentially curing the patient of the disease caused by the genetic defect. Significant academic success has been made with the CRISPR/Cas9 system, though all the work has been completed in animal models. The first in-human testing is to be completed by Editas and its competing commercial entities. Street's take: Overall, we believe consensus recognizes that CRISPR is a significant new technology that could have a large impact on the treatment of genetic diseases. We believe most investors understand the significant potential of CRISPR and the debate centers around how effectively it can be implemented, if the technology is ready to be used in humans and if it is derisked enough yet for an equity investment As is to be expected, some investors see near-term potential while others would prefer to wait for greater clinical experience. Our take: We believe CRISPR is likely to play a major role in gene repair and modification long-term. While we are not entirely sure how long and what the path to clear therapeutic effect will involve, we do believe CRISPR as a technology is proven from a technical perspective and now needs to be translated into humans. Thus, we see great promise in the platform and believe CRISPR has the best chance to create durable, functional cures of many genetic diseases. Gene therapy vs Gene editing? Gene therapy differs from gene editing in both the approach and the possible results seen. Gene therapy generally involves the addition of a new gene to a genome. These new genes can be to replace a defective gene or to add a new gene. The original genes remain intact. Gene editing involves changing the genome, which can involve the addition, deletion, or substitution of components of the genome. CRISPR/Cas9 is not a genetherapy and is a gene editing platform. What is CRISPR/Cas9? CRISPR/Cas9 is a novel method of genetic engineering that uses guide RNA to edit DNA by allowing the enzyme Cas9 to cut and insert the desired genome sequences. The CRISPR/Cas system is formed from CRISPR arrays and CAS genes. CRISPR arrays are clustered, regularly interspaced short palindromic repeats found in certain prokaryotic genomes. CRISPR arrays involve sections of repeated base pair sequence separated by non-repeating spacer sequences. Both the repeated sequences and the spaces come in a large number of variations. Many CRISPR-genes also contain domains associated with DNA manipulation and many of the spacer sequences contained plasm id or phage-derived DNA. CAS, or CRISPR-associated genes, are almost always found adjacent to the repeat arrays. CAS genes come in a variety of subtypes. The CAS9 subtype gene encodes the RNA-guided endonuclease Cas9 which can cleave double stranded DNA. CRISPR RNA (crRNAs) guides with Cas9 proteins and can lead to cleavage of specific sites on DNA and the introduction of templates for gene insertion, as demonstrated in the illustration below. 9 EFTA01100269 Morgan Stanley Editas Medicine] February 29. 2016 MORGAN STANLEY RESEARCH Exhibit 8: Illustration of Gene Editing Using CRISPR/Cas9 DNA target Cas9 CRISPR/Cas9 has the advantage of being cheaper, faster, and easier to use than other techniques. The system is easy to configure between species and allows for an increased precision of insertion. Overall the CRISPR/Cas9 system allows for greater control than usual gene augmentation from gene therapy. The CRISPR/Cas9 technology has seen rapid uptake in a number of academic laboratory groups due to the relative low cost and ease of use. The technology was initially designed for prokaryotic cells but was later adapted to eukaryotic cells, and has since been demonstrated to work in Mammalian cells. A useful review of the history of CRISPR by gado RNA Lander can be found in Celt 164. January 14, 2016. Source: Company Data What other approaches are available for gene editing? Zinc finger nucleases — These are restriction enzymes designed to cleave specific target sequences on DNA. They involve a zinc finger DNA binding domain with site specificity and a DNA-cleavage domain for cutting the phosphodiester bone between nucleic acids in DNA strands. By taking advantage of a cell's intrinsic DNA repair mechanisms, zinc finger nucleases can be used for gene editing. Zinc finger nucleases are being commercialized by Sangamo Biosciences (not covered). Exhibit 9: Illustration of Gene Editing Using Zinc Finger Nucleases (Codon Identification) 5' -ACAAGGAGAGATTICAA TTGAAGAAGTGAAAGA11-3' 3' -1VTICCTCTCTAAAGIT AAA -5' Source: [tont. Physool.. 11 April 1010, 'Emerging gene editing strategies for Duchenne muscular dystrophy targeting stern cells: Carmen leflant. Department of Neurology. David Geffen School of Medicine. Unriersity of California Los Angeles Engineered meganucleases — Meganucleases are naturally occurring proteins that can recognize and cleave specific DNA sequences. They have a DNA recognition sequence that can be modified, thus allowing for engineered meganucleases that can target specific DNA sequences of genetic disorders. Exhibit 10: Illustration of Gene Editing Using Meganucleases Source: [tont. Physeal . 11 April 1014, "Emerging gene editing strategies for Duchenne muscular dystrophy targeting stern cells; Carmen Defiant. Department of Neurology. David Geffen School of Medicine. Un iwesity of California Los Angeles Transcription -activator like effector nucleases (TALENs) — Transcription activator-like effectors (TALEs) are proteins that bind promoter genes and can enhance gene expression. TALEs have a central repeat domain that confers the ability of TALEs to recognize specific DNA sequences along genomes. TALENs are engineered proteins that involve fusing a TALE and a DNA cleavage domain. 10 EFTA01100270 Morgan Stanley Editas Medicine I February 29, 2016 MORGAN STANLEY RESEARCH Exhibit 11: Illustration of Gene Editing Using TALENs (Nucleotide Identification) III I 0 111111111111 5' -ACAAGGAGAGATTTCAA YEGAAGAAGTGAAAGAA- ' Source: Front. Physiol.. 21 April 2014, 'Emerging gene editing strategies for Duchenne muscular dystrophy targeting stem cells,' Carmen Berton*: Department of Neurology. David Geffen School of Medicine. Un ',many of California Los Angeles What is Editas's approach to gene editing? Editas's approach to gene editing has four components:1) nuclease engineering; 2) delivery; 3) control and specificity; and 4) directed editing. Each of these components can be independently optimized in order to develop the best therapeutic candidate. We see Editas as using a comprehensive analytical approach to systematically build a library of CRISPR/Cas9 for a variety of editing approaches. We outline the key features of Editas* approach below. 1. Nuclease Engineering - The CRISPR/Cas9 system involves both a Cas9 protein and an RNA guide molecule. Both of these components can be tailored to allow Editas' gene editing platform to comprehensively target a wide variety of diseases. Editas is developing Cas9 variants tailored to specific genetic defects, and is also making targeted gRNA chemical and structural modifications to build a library of guide RNAs that will allow for enhanced targeting. Additionally, Editas is also using iterative in silico design and high-throughput screening to identify optimal Cas9 / guide RNA combinations. This approach could allow Editas to perform gene editing through various methodologies to address most disease-causing mutations. Exhibit 12: Approaches Available to Gene Editing for Editas CUT & REVISE CUT & REMOVE TGCACCTGAArH t TGCACCTGAAXXXGGCA "I irzi.-4 TGAATGGrA ,— _ 4 IS° TGGCA CUT & REPLACE CCTGAATGCCA 4 -V TAGTCGCATCCCGCA Source: Company Data 2. Delivery - Different disease indications will involve different cell types and tissue structures and thus delivery mechanisms will need to be specifically designed for each disease type. Editas's CRISRP/Cas9 system is adaptable to different delivery modalities, and the company intends to use existing delivery technologies for in vivo and ex vivo delivery through modalities including viral vectors, nanoparticles, and electroporation. Preclinical data for both ex vivo and in vivo genome editing has provided encouraging early results, as demonstrated below. Next-generation delivery methods will be studied based on product needs going forward. 3. Control & Specificity — One of the largest advantages of gene editing techniques over conventional gene therapy techniques is the high level of specificity available when altering, deleting, or inserting new genetic material. The specific DNA cut sites must be optimized in order to deliver optimal therapy. Cellular exposure to the Cas9-guide/RNA complex can also affect the outcome and can be optimized to provide maximal benefit. Editas is exploring codon optimization for Cas9, to be able to identify the ideal codon set for each tissue. The company is also working to identify tissue-specific promoters for both guide RNAs and Cas9 proteins. Specific methodologies the company is exploring to control the editing process include self-targeting gRNAs to turn expression off, developing small molecule modulators of Cas9, and incorporating functional motifs and 11 EFTA01100271 Morgan Stanley Editas Medicine I February 29. 2016 MORGAN STANLEY RESEARCH Exhibit 14: Ex vivo: mRNA & RNP Delivery of PD- Exhibit 13: In viva Factor VII Gene Knockdown in 1 Targeted Cas9 in T-Cells the Liver with AAV so 40 00c 30 -o 20 10 0 Source: Company Data d Day 2 Day 3 Untreated harkat developing variants to control cellular Cas9 degradation. 4. Directed editing — There are several mechanisms by which the CRISPR/Cas9 system and the cell's intrinsic repair systems can work together to cut and repair the target cell's DNA. Non-homologous end joining (NHEJ) Source Company Data is a method of DNA repair that occurs in the absence of a DNA template for each cell to copy that leads to small insertions and deletions. This is the best option when treatment requires the deletion of gene segments. The difficulty of using this mechanism increases as the length of DNA requiring deletion increases. Homologous directed repair (HDR) is a method of DNA repair that occurs in the presence of a DNA template. This method leads to the replacement of defective sequences with functional ones. This technique is the best option for inserting new genes. Editas is studying the impact that Cas9 variants and different cutting approaches have on the NHEJ and HDR repair pathways, and is working to enhance the efficiency of HDR by using donor DNA modifications. Important Papers in CRISPR/Cas9's development The development of CRISPR/Cas9 for Human Cells "Multiplex Genome Engineering Using CRISPR/Cas Systems" - L Cong, et al. In Science 2013 This paper demonstrates that RNA-guided Cas9 nucleases can be engineered that precisely target genomic loci in mammalian cells and cause double stranded breaks in mammalian chromosomes. "RNA-Guided Human Genome Engineering via Cas9" - P Mali, et al. In Science 2013 This paper focuses on the development and testing of engineered CRISPR/Cas9 with a custom guide RNA in human cells. This paper established that this technology was capable of human genome engineering. These tests were done in vitro. Improving Delivery "Cationic lipid-mediated delivery of proteins enables efficient protein-based genome editing in vitro and in vivo" Zuris, et al. In Nature Biotechnology 2014 This paper discusses a delivery system that can be applied to the CRISPR/Cas9 system proteins. Experiments were performed both in vitro and in vivo that demonstrated that the Cas9 system could be delivered and would 12 EFTA01100272 Morgan Stanley modify the target genome. Editas Medicine] February 29. 2016 MORGAN STANLEY RESEARCH "In vivo genome editing using Staphylococcus aureus Cas9" - F Ran et al. In Nature 2015 Earlier studies of Cas9 mainly relied on Streptococcus pyogenes derived Cas9s (SpCas9). These are larger proteins and limit the useful applications. This paper demonstrated that the significantly smaller Cas9 from Staphylococcus aureus can achieve similar efficacies to SpCas9, while maintaining specificity and efficacy. Improving Characterization "GUIDE-seq enables genome-wide profiling of off-target cleavage by CRISPR-Cas nucleases" - S Tsai et al. In Nature Biotechnology 2014 This paper highlights a detection system for identifying off-target DNA double strand breaks caused by CRISPR/Cas9 nucleases. Experiments with 13 CRISPR systems in 2 human cell lines demonstrated a high level of variability in off-target activities that were not detected by other computational and experimental detection methods. It was also shown that truncated RNA guides led to a reduction in off target double strand breaks. in vivo interrogation of gene function in the mammalian brain using CRISPR-Cas9" - L Swiech et al. In Nature Biotechnology 2014 In this study an AAV vector was used to deliver CRISPR/Cas9 derived from Streptococcus pyogenes to adult mouse brain cells targeting specific genes. The genome editing resulted in biochemical, genetic, electrophysical, and behavioral changes. Improving Specificity "Double Nicking by RNA-Guided CRISPR Cas9 for Enhanced Genome Editing Specificity" - F Ran et al. In Cell 2013 This paper discusses the use of paired RNA guides to introduce double strand breaks into target DNA. By using two guides with known spacing the specificity of the DNA cuts drastically increases. These experiments were performed in mouse zygotes. "Improving CRISPR-Cas nuclease specificity using truncated guide RNASs" - Y Fu et al. In Nature Biotechnology 2014 In this paper, researchers compare guide RNAs with truncated guide RNAs in terms of off-target effects. They demonstrate that the truncated guide RNAs have as much as a 5000-fold decrease in undesired mutagenesis without affecting on-target editing. "Fusion of catalytically inactive Cas9 to Fokl nuclease improves the specificity of genome modification" -.I Guilinger et al. In Nature Biotechnology 2014 The fusion of inactive Cas9 and Fokl nuclease improves DNA cleavage specificity. These fused Cas9 complexes had a significantly higher specificity than WT Cas9 proteins. "Dimeric CRISPR RNA-guided Fokl nucleases for highly specific genome editing" - S Tsai et al. In Nature Biotechnology 2014 This paper describes dimeric RNA-guided Fokl nucleases that have high efficacy and specificity in their DNA targets by recognizing extended genetic sequences. This system uses two guide RNAs with fixed and known spacing to ensure specificity. "Engineered CRISPR-Cas9 nucleases with altered PAM specificities" - B Kleinstiver et al. In Nature 2015 13 EFTA01100273 Morgan Stanley Editas Medicine I February 29. 2016 MORGAN STANLEY RESEARCH The range of sequences that can be identified with a Cas9 protein is determined by the specific protospacer adjacent motifs (PAM) on the targeted DNA. This paper demonstrates that the specificity of the CRISPR/Cas9 system for PAMs can be altered so that a wider range of targets is available to therapy. "High Fidelity CRISPR-Cas9 nucleases with no detectable genome-wide off target effects" - B Kleinstiver et al. In Nature 2015 This paper describes a variant of SpCas9 that minimized non-specific DNA contacts that results in the retention of on-target activity while rendering nearly all off target effects undetectable. "Rationally engineered Cas9 nucleases with improved specificity - I Slaymaker et al in Science 2015 In this paper, structure-guided protein engineering is used to alter and improve the specificity of SpCas9. These altered Cas9 proteins reduced off-target effects while still maintaining on-target editing. Key Academic Institutions with CRISPR/Cas9 Progams The development of CRISPR/Cas9 technology provides a model for what can be done through incremental progress in science. In the late 1980s/early 1990s, Dr. Francisco Mojica, a post doctoral student at the University of Alicante in Spain, noticed a unique/repeating genetic sequence in Haloferax mediterranei. It altered between a repeat sequence of 30 base pairs followed by -36 spacer base nucleotides. Dr. Mojica initially called the finding short regularly spaced repeats (SRSRs), eventually renaming it to clustered regularly interspaced palindromic repeats or CRISPR. This finding from a microbe usually found in salty environments, and noted for its ability to tolerate excessively high salinity, marks the beginning of an incremental set of advances that eventually led to the gene editing technologies being pursued by Editas and its competitors. Post the early discovery in Spain, there was a great deal of work that needed to be done to determine the purpose (if any) of these repeat sequences. After years of work, Dr. Mojica noted that these spacers were noted in E. coli strains with bacterial resistance leading to the hypothesis that CRISPR's purpose was part of the adaptive immune system offering a way for cells to adapt to their environment. While Dr. Mojica made some early findings, many others contributed new discoveries at a rapid pace. From Horvavth, Barrangou and Moineau who discovered that Cas nucleases contributed to this adaptive response to Marrafani and Sontheimer, who discovered that CRISPR specifically targeted DNA. As more incremental advances occurred it became increasingly apparent that not only is CRISPR, and in particular the Cas9 nucleotide, an important component of the adaptive immune system but also that it could be utilized as a gene editing tool. Such a tool would have immense potential for treating a massive array of diseases and disorders. Below we outline the key academic institutions who claim to have IP for CRISPR/Cas9 and their corporate affiliation. Broad Institute, Massachusetts Institute of Technology, Harvard University • Contribution: Dr. Feng Zhang, one of the founders of Editas and a principal scientist in the intellectual property dispute, conducted his research at the Broad Institute/MIT. Zhang's team examined CRISPR-Cas9 and in partnership with Dr. George Church from Harvard University learned that the nuclease Cas9 could be used to cut DNA in a highly specific location. This work was published in January 2013 in the journal Science. Dr. Zhang's team demonstrated that the cut DNA was replaceable with another piece of DNA, changing the overall sequence of the gene. Zhang and team also discovered Cpfl, which could also have similar application with the advantage of being a smaller protein than Cas9. Dr. George Church of Harvard University, and also a Broad Institute researcher, conducted work concurrently with Dr. Zhang and reported similar findings. Dr. Church's work was also published in the January 2013 issue of Science. It is important to note that both Church and Zhang also were able to translate the use of CRISPR/Cas9 into mammalian cells. ■ Intellectual Property Status: The Broad Institute filed for patents through the Prioritized Patent Examination Program, a fast-track review program. In December 2015, only six months after application, the US Patent and Trademark Office (USPTO) issued 23 patents 14 EFTA01100274 Morgan Stanley Editas Medicine] February 29. 2016 MORGAN STANLEY RESEARCH with respect to CRISPR/Cas9, 13 of which were issued to the Broad Institute, MIT and Dr. Feng Zhang. Harvard University was issued 4 patents relating to CRISPR. ■ Affiliation: Editas Therapeutics University of California, Berkeley, University of Vienna • Contribution: Dr. Jennifer Doudna of the University of California at Berkeley partnered with Dr. Emmanuelle Charpentier from the University of Vienna. Much of their work together was focused on Cas9. They made similar discoveries as Dr. Zhang and Dr. Church's team, demonstrating that the enzyme could be used to cut DNA. Dr. Doudna was one of the original founders of Editas; although, she left the company to start Caribou Biosciences, a company that eventually cofounded Intellia Therapeutics. ■ Intellectual Property Status: Doudna and Charpentier submitted a patent application for the UC Berkeley and the University of Vienna to the USPTO seven months ahead of Zhang. However, the application was submitted via the normal track as opposed to the fast tracked Zhang application. As such, the Broad Institute teams were granted broad Cas9 patents. UC Berkeley & University of Vienna initiated multiple Suggestions of Interference proceedings to challenge the Broad Institute patents. • Affiliation: Intellia Therapeutics and CRISPR Therapeutics are affiliated with Dr. Doudna and Dr. Charpentier, respectively. University of Vilnius ■ Contribution: Dr. Virginijus Siksnys of the University of Vilnius also concurrently conducted and published work evaluating Cas9 activity. In this work, it was demonstrated that Cas9 could be targeted to cut double stranded DNA exactly three nucleotides from the protospacer adjacent motif (PAM) sequence. Dr. Siksnys submitted his work for publication, a month before the Doudna/Charpentier paper had been published (after it was fast-tracked through Science's review process) • Intellectual Property Status: Dr. Siksnys filed patent application in March of 2012 based on this work. • Affiliation: Dupont for agricultural purposes. The Rockefeller University • Contribution: Dr. Luciano Marraffini is affiliated with the Rockefeller University. In his work he determined that CRISPR specifically targets DNA. In his work, he concluded that CRISPR was a programmable restriction enzyme that could potentially be employed as a gene editing technology. ■ Intellectual Propery Status: The Rockefeller University is a joint patent applicant on certain patent applications along with the Broad Institute. • Affiliation: Intellia Therapeutics ToolGen, Inc. • Contribution: Affiliated with Seoul National University in South Korea, ToolGen has developed genetic tools based on zinc finger engineering technology, with technology ultimately evolving toward the use of Cas9 nucleases. The company has also refined a process called double-nicking approach with zinc finger nucleases to clip the DNA. The belief 15 EFTA01100275 Morgan Stanley Editas Medicine I February 29. 2016 MORGAN STANLEY RESEARCH is that this will reduce off-target modification. • Intellectual Propery Status: ToolGen has filed Suggestions of Interference claims that two of their patents interfere with five Broad Institute patents. UC Berkeley has filed a similar claim on these five Broad Institute patents. • Affiliation: Thermo Fisher Scientific Exhibit 15: Various Academic Institutions and Players With CRISPR/Cas9 Programs Academic institution Contribution iP Status /41/6.0.1 Massachuswts Institute of Technology (MIT) - Broad Institute Harvard University- Broad Institute Dr. Zhang lad one effort to develop Cas9 for the purpose of gene editing. Granted initial Patents in large Part due TO decision to apt* through USPTO Prioritized Patent F commotion Program, leading to 6-month review process. Dr. Church of Harvard and Broad Institute conducted work on Cas9 concurrently with Dr. 2hang. First to be granted Patents i /X CRISP'', Cas9 gene editing technology m December 2015 Granted multiple initial patents in tandem WW1 MIT - Broad Institute. (Odes (dila, UC Berkeley ntellia University of Vienna Dr. Doudna conducted similar work on Cass for gene editing in unrelated effort. An original founder of f ditas, left to form Caribou Biosciences which co-founded Intellia. Dr. Charpentret worked in partnership with Dr. Doudna in Cas9 research. She is a founder of CRISPR Therapeutics Applied for Cas9 patents 7 months ahead of thong's team However, fell behind because Zhang's fast-tracked process. Filed multiple Suggestions of rnterference claims. Univensty of Vienna was named as part of VC Berkety's patent application and is included in Suggestion of Interference claims. Intelba University of Vilnius Dr. Siltsnys of the Unwersity of Vilnius showed that Cas9 could be targeted to cut double stranded DNA exactly three nucleotides from the protospacer adjacent motif (PAM) sequence. His work was submitted for publication l'month before Doudnagharpentier's efforts were published. Filed for patents in 2012. Dupont The Rockefeller University Dr. Marrahni conducted much work on CRISPR, being one of the first to determine it actually targeted DNA and suggesting it could be a gene editing technology Rockefeller is named on five Broad Institute Patents. Though they received nO rights as part of Broad's license agreement with Editas. 'MOM ToolGen, Inc. A Korean company affiliated with Seoul National University that used zinc finger technology in CRISPR research, moving eventually toward Cas9. Employs double-nicking to potentially reduce off target effects rnitiated Suggestion of Interference Proceeding regarding five Broad Institute patents Thermo Fisher Scientific Source: Cell 164, January 14, 2016 and Morgan Stanley 16 EFTA01100276 Morgan Stanley Editas Medicine] February 29. 2016 MORGAN STANLEY RESEARCH Debate 2 - What Is Going On With IP And What Potentially Can be a Plausible Outcome? Overview: In early January, an interference proceeding was declared between the Broad Institute/Harvard/MIT (from whom Editas has licensed its IP) and University of California, University of Vienna and Emmanuelle Charpentier (from whom Intellia and Crispr Therapeutics have licensed IP). At the core of the debate is who was first to invent the use of CRISPR/Cas9 in eukaryotic cells (particularly mammalian cells). The Broad has 12 issued patents named in the interference proceeding versus a yet to be issued University of California patent. Because California claims to have invented first it has been named the senior party while the Broad the junior party, though those designations could change. The total interference proceeding is expected to play out over the next two years. Potential outcomes could be that the Patent Trial and Appeal Board (PTAB) invalidates the Broad's IP, leaving Editas without freedom to operate, that the PTAB does not award a patent to the University of California, meaning Editas has the only granted IP for CRISPR/Cas9, or a limiting of claims for both parties with issued IP for both sides. An appeal is also possible after the initial ruling. Street's take: Investors have seen a few interference proceedings before, namely, Gilead versus Idenix for sofosbuvir and Biogen versus Forward Pharma related to Tecfidera IP. However, neither case has been potentially so central to the investment debate on the stock. Thus, for many investors this is an area they find hard to completely derisk. Despite that, we believe the prevailing opinion is either to invest broadly across the CRISPR/Cas9 space, thereby having investments in both parties or assuming, as do we, that given the breadth of the Editas IP, while some IP may be narrowed, the likelihood that all IP will fall is low. Our take: We believe a plausible outcome of the interference could be that both parties would end up with issued IP for CRISPR/Cas9, requiring both parties to engag