10-K

UNITED STATES

SECURITIES AND EXCHANGE COMMISSION

Washington, D.C. 20549

FORM 10-K

(Mark One)

For the fiscal year ended December 31, 2025

Commission File Number 001-40522

Monte Rosa Therapeutics, Inc.

(Exact name of Registrant as specified in its Charter)

321 Harrison Avenue, Suite 900Boston, Massachusetts 02118

(Address of principal executive offices) (Zip Code)

Registrant’s telephone number, including area code: (617) 949-2643

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

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

Common stock, par value $0.0001 per share GLUE The 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. ☐

If securities are registered pursuant to Section 12(b) of the Act, indicate by check mark whether the financial statements of the registrant included in the filing reflect the correction of an error to previously issued financial statements. ☐

Indicate by check mark whether any of those error corrections are restatements that required a recovery analysis of incentive-based compensation received by any of the registrant’s executive officers during the relevant recovery period pursuant to §240.10D-1(b). ☐

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, based on the closing price of the shares of common stock on NASDAQ on June 30, 2025, was $212 million.

The number of shares of Registrant’s Common Stock outstanding as of March 2, 2026, was 80,015,667.

DOCUMENTS INCORPORATED BY REFERENCE

Portions of registrant's definitive proxy statement for its annual meeting of shareholders to be filed within 120 days after the close of the registrant's fiscal year are incorporated by reference to into Part III of this annual report on Form 10-K.

Table of Contents

Page

PART I

Item 1. Business 1

Item 1A. Risk Factors 85

Item 1B. Unresolved Staff Comments 141

Item 1C. Cybersecurity 141

Item 2. Properties 141

Item 3. Legal Proceedings 142

Item 4. Mine Safety Disclosures 142

PART II

Item 6. [Reserved] 144

Item 7A. Quantitative and Qualitative Disclosures About Market Risk 157

Item 8. Financial Statements and Supplementary Data 157

Item 9A. Controls and Procedures 158

Item 9B. Other Information 158

Item 9C. Disclosure Regarding Foreign Jurisdictions that Prevent Inspections 159

PART III

Item 10. Directors, Executive Officers and Corporate Governance 160

Item 11. Executive Compensation 160

Item 14. Principal Accountant Fees and Services 160

PART IV

Item 15. Exhibits and Financial Statement Schedules 161

SPECIAL NOTE REGARDING FORWARD LOOKING STATEMENTS

This Annual Report on Form 10-K, or Annual Report, contains forward-looking statements which are made pursuant to the safe harbor provisions of Section 27A of the Securities Act of 1933, as amended, or the or the Securities Act, and Section 21E of the Securities Exchange Act of 1934, as amended, or the Exchange Act. All statements other than statements of historical facts contained in this Annual Report are forward-looking statements. In some cases, you can identify forward-looking statements by terminology such as “may”, “will”, “should”, “expects”, “intends”, “plans”, “anticipates”, “believes”, “estimates”, “predicts”, “potential”, “continue” or the negative of these terms or other comparable terminology. These statements are not guarantees of future results or performance and involve substantial risks and uncertainties. Forward-looking statements in this Annual Report include, but are not limited to, statements about:

•

the initiation, timing, progress, results, costs, and any expectations and/or predictions of success of our current and future research and development programs and preclinical studies, including our expectations for our molecular glue degraders, or MGDs, molecules, including our GSPT1-directed MGD MRT-2359, our VAV1-directed MGD MRT-6160, our NEK7-directed MGDs, including MRT-8102, and our CDK2 and CCNE1 MGDs;

•

the initiation, timing, progress, results, costs, and any expectations and/or predictions of success of our current and any future clinical trials, including our clinical trials for our GSPT1-directed MGD MRT-2359, our NEK7 directed MGD MRT-8102, and for our out-licensed VAV1 directed MGD MRT-6160, including statements regarding the nature of or the timing for when any results of any clinical trials will become available;

•

our ability to continue to develop our proprietary discovery engine, called QuEENTM, and to expand our proteomics and translational medicine capabilities;

•

the potential advantages of our discovery engine technology and product candidates;

•

the extent to which our scientific approach and discovery engine technology may target proteins that have been considered undruggable or inadequately drugged;

•

our plans to submit Investigational New Drug, or IND, applications to the U.S. Food and Drug Administration, or the FDA, for current and future product candidates;

•

the potential benefits of strategic collaborations and our ability to enter into strategic collaborations with third parties who have the expertise to enable us to further develop our biological targets, product candidates and discovery engine technologies, including our agreements with Novartis AG, or Novartis, for MRT-6160 and other discovery programs and our agreement with F. Hoffmann-La Roche Ltd., or Roche Basel, and Hoffmann-La Roche Inc., or Roche US, and together with Roche Basel referred herein as Roche;

•

our ability to obtain and maintain regulatory approval of our product candidates;

•

our ability to maintain and expand, including through third-party vendors, our library of MGDs;

•

our ability to manufacture, including through third-party manufacturers, our product candidates for preclinical use, future clinical trials and commercial use, if approved;

•

our ability to commercialize our product candidates, including our ability to establish sales, marketing and distribution capabilities for our product candidates;

•

the rate and degree of market acceptance of our product candidates;

•

the size and growth potential of the markets for our product candidates, and our ability to serve those markets;

•

our ability to establish and maintain intellectual property rights covering our current and future product candidates and technologies;

•

the implementation of our business model and strategic plans for our business, product candidates, and technology;

•

estimates of our future expenses, revenues, capital requirements, and our needs for additional capital;

•

our expected use of proceeds from sales of our common stock in "at-the-market" offerings and other offerings, and the period over which such proceeds, together with existing cash, will be sufficient to meet our operating needs;

•

our ability to obtain funding for our operations necessary to complete further development and commercialization of our product candidates;

•

our financial performance;

•

developments in laws and regulations in the United States, or the U.S., and foreign countries;

•

the success of competing therapies that are or may become available;

•

our ability to attract and retain key scientific or management personnel;

•

the effect of global economic uncertainty and financial market volatility caused by economic effects of rising inflation and interest rates, global health crises, geopolitical events, elections, changes in international trade relationships and military conflicts on any of the foregoing or other aspects of our business or operations;

•

the effect of any geopolitical conflicts or new or increased international tariffs, including mitigation efforts and economic effects, on any of the foregoing or other aspects of our business operations, including but not limited to our preclinical studies, ongoing clinical trials and future clinical trials; and

•

other risks and uncertainties, including those listed under Item 1A, “Risk Factors.”

Any forward-looking statements in this Annual Report reflect our current views with respect to future events and with respect to our future financial performance, and involve known and unknown risks, uncertainties and other factors that may cause our actual results, performance, or achievements to be materially different from any future results, performance, or achievements expressed or implied by these forward-looking statements. Factors that may cause actual results to differ materially from current expectations include, among other things, those described under Item 1A, “Risk Factors” and elsewhere in this Annual Report. Given these uncertainties, you should not place undue reliance on these forward-looking statements.

All of our forward-looking statements are as of the date of this Annual Report only. In each case, actual results may differ materially from such forward-looking information. We can give no assurance that such expectations or forward-looking statements will prove to be correct. An occurrence of or any material adverse change in one or more of the risk factors or risks and uncertainties referred to in this Annual Report or included in our other public disclosures or our other periodic reports or other documents or filings filed with or furnished to the Securities and Exchange Commission, or the SEC, could materially and adversely affect our business, prospects, financial condition and results of operations. Except as required by law, we do not undertake or plan to update or revise any such forward-looking statements to reflect actual results, changes in plans, assumptions, estimates or projections or other circumstances affecting such forward-looking statements occurring after the date of this Annual Report, even if such results, changes or circumstances make it clear that any forward-looking information will not be realized. Any public statements or disclosures by us following this Annual Report that modify or impact any of the forward-looking statements contained in this Annual Report will be deemed to modify or supersede such statements in this Annual Report.

We may from time to time provide estimates, projections and other information concerning our industry, the general business environment, and the markets for certain diseases, including estimates regarding the potential size of those markets and the estimated incidence and prevalence of certain medical conditions. Information that is based on estimates, forecasts, projections, market research or similar methodologies is inherently subject to uncertainties, and actual events, circumstances or numbers, including actual disease prevalence rates and market size, may differ materially from the information reflected in this Annual Report. Unless otherwise expressly stated, we obtained this industry, business information, market data, prevalence information and other data from reports, research surveys, studies and similar data prepared by market research firms and other third parties, industry, medical and general publications, government data, and similar sources, in some cases applying our own assumptions and analysis that may, in the future, prove not to have been accurate.

TRADEMARKS

Solely for convenience, our trademarks and trade names in this report are sometimes referred to without the ® and TM symbols, but such references should not be construed as any indicator that we will not assert, to the fullest extent under applicable law, our rights thereto.

iii

SUMMARY OF RISK FACTORS ASSOCIATED WITH OUR BUSINESS

Our business is subject to numerous material and other risks and uncertainties that you should be aware of in evaluating our business. These risks are described more fully in Part II, “Item 1A—Risk Factors,” and include, but are not limited to, the following:

•

We are a biotechnology company with a limited operating history and have not generated any revenue to date from drug sales and may never become profitable.

•

We have incurred significant operating losses since our inception and anticipate that we will incur continued losses for the foreseeable future.

•

We are very early in our development efforts. Several of our programs are still in the preclinical stages of drug development. If we are unable to commercialize our product candidates or experience significant delays in doing so, our business will be materially harmed.

•

Our approach to the discovery and development of product candidates, which may also be referred to herein as development candidates, based on our QuEENTM discovery engine is novel, which makes it difficult to predict the time, cost of development and likelihood of successfully developing any product candidates.

•

We will need to raise substantial additional funding before we can expect to complete development of any of our product candidates or generate any revenues from product sales.

•

We may not be successful in our efforts to identify or discover additional product candidates or we may expend our limited resources to pursue a particular product candidate or indication and fail to capitalize on product candidates or indications that may be more profitable or for which there is a greater likelihood of success.

•

If we are unable to successfully develop our current programs into a portfolio of product candidates, or experience significant delays in doing so, we may not realize the full commercial potential of our current and future product candidates.

•

If we encounter difficulties enrolling patients in our clinical trials, these clinical development activities could be delayed or otherwise adversely affected.

•

If we are unable to advance our product candidates through clinical development, obtain regulatory approval and ultimately commercialize our product candidates, or experience significant delays in doing so, our business will be materially harmed.

•

Even if we receive marketing authorization for our product candidates, we will be subject to extensive ongoing regulatory obligations and continued regulatory review, which may result in significant additional expense and we may be subject to penalties if we fail to comply with regulatory requirements or experience unanticipated problems with our product candidates.

•

If we are unable to obtain and maintain patent and other intellectual property protection for our technology and product candidates or if the scope of the intellectual property protection obtained is not sufficiently broad, our competitors could develop and commercialize technology and drugs similar or identical to ours, and our ability to successfully commercialize our technology and drugs may be impaired, and we may not be able to compete effectively in our market.

•

Our future success depends on our ability to retain key executives and to attract, retain and motivate qualified personnel.

•

Business disruptions could seriously harm our future revenue and financial condition and increase our costs and expenses.

•

Our executive officers, directors, principal stockholders and their affiliates exercise significant influence over our company, which will limit our stockholders' ability to influence corporate matters and could delay or prevent a change in corporate control.

PART I

Item 1. Business

Overview

We are a clinical-stage biotechnology company developing a portfolio of novel and proprietary molecular glue degraders, or MGDs. MGDs are small molecule drugs that employ the body’s natural protein destruction mechanisms to selectively degrade therapeutically relevant proteins, in effect editing the human proteome. MGDs function by inducing the engagement of an E3 ligase, such as cereblon, with defined structural features on surfaces of target proteins. These target proteins are also referred to as neosubstrates. The E3 ligase then tags the target protein for degradation by adding a molecular mark known as ubiquitin. We believe our MGDs provide significant advantages over existing therapeutic modalities, including other protein degradation approaches.

We have developed a proprietary and industry leading discovery engine, called QuEENTM (an abbreviation for “Quantitative and Engineered Elimination of Neosubstrates”) to enable our unique target-centric MGD discovery and development approach and our rational design of MGD product candidates.

To date, our QuEENTM discovery engine has identified numerous proteins for potential targeting by our MGDs, including those targeted by product candidates in our pipeline. We combine our artificial intelligence or “AI” / machine learning or “ML” engines with multiple proprietary experimental tools to identify therapeutically relevant target proteins amenable to degradation by our MGDs. We are continuously increasing our understanding of how MGDs function and we are using this understanding to develop design principles for the engineering of new MGDs. This growing expertise manifests in our expanding MGD library as well as our discovery and development pipeline. Using our insights, knowhow, and technology platform we have generated a library of MGDs that forms the basis for our MGD programs. At present, our library comprises a diverse set of rationally designed small molecules representing more than 1000 unique low molecular weight scaffolds and over 75,000 different MGD molecules. We also use our insights and learnings to continuously update and improve QuEENTM and our MGD library, consistently increasing the power of the discovery engine.

We prioritize our product development to address therapeutic targets backed by strong biological and genetic rationales. We are focused on developing solutions to clinically important indications, including indications in immunology, inflammation, cardiology, oncology, and others. To date, our discovery engine has resulted in three programs in clinical development: MRT-6160, a VAV1-directed MGD for immune-mediated diseases; MRT-8102, a NEK7-directed MGD for inflammatory diseases driven by IL-1β, IL-6, and the NLRP3 inflammasome; and MRT-2359, a GSPT1-directed MGD for metastatic castration resistant prostate cancer (mCRPC).

MRT-6160 is a VAV1-directed MGD being developed for immune-mediated diseases. Our preclinical studies showed that targeted degradation of VAV1 protein via an MGD modulates both T- and B-cell receptor activity. Our VAV1 MGD, MRT-6160, showed promising activity in preclinical models of neurologic and systemic autoimmune and inflammatory diseases and thus we believe has the potential to provide therapeutic benefit in multiple immune-mediated diseases, such as inflammatory bowel disease, rheumatoid arthritis, dermatological disorders, and multiple sclerosis.

In October 2024, we announced a global exclusive development and commercialization license agreement with Novartis under which we granted to Novartis an exclusive license to develop, manufacture, and commercialize VAV1-directed MGDs including MRT-6160, starting with Phase 2 clinical studies. We received from Novartis an upfront payment of $150 million and are eligible to receive up to $2.1 billion in development, regulatory, and sales milestones, beginning upon initiation of Phase 2 studies and including potential development and regulatory milestone payments, exceeding $1.5 billion if multiple indications achieve regulatory approval in multiple territories. We and Novartis also agreed to a net profit and loss sharing arrangement, in which we will co-fund any global clinical development from Phase 3 onwards and will share 30% of any profits and losses associated with the manufacturing and commercialization of the licensed products in the United States. We are eligible to receive from Novartis potential sales milestone payments and tiered royalties in connection with sales outside of the United States. We were responsible for costs associated with the now completed Phase 1 clinical study and Novartis will be responsible for costs associated with any subsequent clinical studies except for the Phase 3 costs covered by us under the profit and loss sharing agreement.

In March 2025, we announced initial clinical results from our Phase 1 study of MRT-6160, demonstrating deep VAV1 degradation of greater than 90%, significant T and B cell functional inhibition, including profound inhibition

of cytokine release from T and B cells ex-vivo, and a generally favorable safety and tolerability profile. We believe these data support a clear path to multiple Phase 2 studies and broad potential applications in immune-mediated diseases. We expect that our collaborator, Novartis, will initiate multiple Phase 2 studies of MRT-6160 in immune-mediated diseases in 2026.

MRT-8102 is a NEK7-directed MGD targeting diseases and inflammatory conditions driven by the NLRP3 inflammasome, IL-1, and IL-6. The NLRP3 inflammasome is a multi-protein complex that serves as a central node for integrating signals from pathogens, damage, and stress, and triggers the production of pro-inflammatory cytokines. Aberrant NLRP3 inflammasome activation and the subsequent release of active interleukin-1β (IL-1β) and interleukin-18 (IL-18) have been implicated in multiple inflammation-driven diseases, including atherosclerotic cardiovascular disease (ASCVD), gout, hidradenitis suppurativa, pericarditis, osteoarthritis, and obesity. NEK7, functioning as a scaffolding protein, facilitates assembly and activation of the NLRP3 inflammasome in a kinase-independent manner, suggesting that degradation of NEK7 with an MGD molecule would be a potentially attractive therapeutic approach to preventing NLRP3 inflammasome activation and associated downstream cytokine production.

In January 2026, we announced positive interim data from an ongoing Phase 1 clinical study (now called GFORCE-1) evaluating MRT-8102. In subjects with elevated cardiovascular disease (CVD) risk, MRT-8102 demonstrated rapid and durable reductions in systemic inflammation. After four weeks of MRT-8102 treatment in subjects with elevated CVD risk, C-reactive protein (CRP) levels were reduced by 85%, and 94% of study participants achieved CRP values below 2 mg/L, a threshold associated with reduced CVD risk. The single ascending dose (SAD) and multiple ascending dose (MAD) cohorts demonstrated deep and sustained NEK7 degradation at doses from 5 mg to 400 mg. A favorable safety profile was observed with mild to moderate adverse events (AEs) and no evidence of increased infection risk.

Our ongoing GFORCE-1 Study of MRT-8102 in subjects with elevated CVD risk has been expanded to multiple dose levels to accelerate development in ASCVD. We anticipate results from this study in H2 2026. We plan to initiate a Phase 2 ASCVD study, GFORCE-2, (in elevated CVD risk patients defined by Stage 3/4 chronic kidney disease and elevated CRP) in H2 2026, a Phase 2 study of MRT-8102 in patients with gout flares, GFORCE-3, in Q4 2026 or Q1 2027, and a Phase 2 study in patients with hidradenitis suppurativa, GFORCE-4, in H1 2027. Furthermore, we expect to submit an IND application for a next-generation NEK7-directed MGD in 2026.

MRT-2359 is an orally bioavailable MGD targeting the translation termination factor protein GSPT1 and is currently in clinical development for potential use in MYC-driven tumors, with a focus on metastatic castration-resistant prostate cancer, or mCRPC. GSPT1 (also known as eRF3a) is a translation termination factor that helps catalyze the termination of protein synthesis, facilitating the release of mRNA and newly synthesized protein from the ribosomal protein synthesis machinery. We have identified GSPT1 as a potential therapeutic vulnerability for MYC-driven cancers, including mCRPC. MRT-2359, our GSPT1-directed MGD, was designed to preferentially affect growth and survival of cancer cells addicted to protein translation, such as those driven by high expression and activity of MYC family transcription factors. Our preclinical studies showed that once-daily oral dosing of MRT-2359 led to potent antitumor activity in MYC-driven cell-line- and patient-derived xenograft models, and pointed to mCRPC as a potential indication for MRT-2359.

In December 2025, we announced positive interim data from an ongoing Phase 1/2 clinical study evaluating MRT-2359 in combination with enzalutamide in heavily pretreated patients with metastatic castration-resistant prostate cancer (mCRPC). We provided further updates from this ongoing clinical study in February 2026, including at the ASCO Genitourinary Cancers Symposium held in San Francisco on February 26-28, 2026. In our February 2026 update we showed that in the subset of mCRPC patients with androgen receptor (AR) mutations, treatment with MRT-2359 in combination with enzalutamide led to a 100% PSA response rate in patients identified as having AR mutations (5 out of 5 patients). In addition, MRT-2359 plus enzalutamide demonstrated a 100% disease control rate in this patient subset, per RECIST criteria, including 2 RECIST partial responses and 3 with stable disease. In total, our study included 15 evaluable patients, including the 5 patients identified as having AR mutations. Across those 15 evaluable patients, the overall disease control rate was 67% (10 of 15), with 10 of 15 patients showing tumor size reductions of target lesions, including the 2 RECIST partial response patients that were also identified as having AR mutations.

Based on the success of our reported Phase 1/2 studies in mCRPC patients, we plan to initiate a Phase 2 study of MRT-2359 in combination with a second-generation AR inhibitor in mCRPC patients with AR mutations, with potential to expand the study into additional patient subsets, including patients naive to 2nd generation AR inhibitors. The study is anticipated to start in 2026.

We are also advancing programs directed at cyclin E1 (CCNE1) and cyclin-dependent kinase 2 (CDK2), key drivers of cell cycle progression in cancer.

Cyclin E1 is a protein that plays a crucial role in the cell cycle, and is a frequently amplified non-enzymatic driver oncogene relevant in multiple solid tumors that has not been druggable by conventional modalities. We believe that our proprietary Cyclin E1 MGDs represent a potential novel therapeutic approach for treatment of such solid tumors by directly and selectively targeting Cyclin E1. We believe our cyclin E1 MGDs could provide a highly differentiated alternative and additional approach to other cell-cycle focused therapeutics currently in development.

We expect to submit an IND application for a cyclin E1-directed MGD in 2026.

Our CDK2-directed MGDs have demonstrated superior selectivity for CDK2 in preclinical models as compared to several clinical-stage small molecule CDK2 ATP-site inhibitors, which we believe will be important to mitigate toxicity limitations reported for CDK2 inhibitors in development. In preclinical models of ER+ breast cancer our CDK2 MGDs reduced tumor burden when added to standard of care therapy. We believe our preclinical data supports further clinical evaluation of our CDK2 MGDs as a potential improvement over current standard of care therapies in ER+ breast cancer, and potentially without the toxicity limitations reported for CDK2 inhibitors currently in development.

Our proprietary QuEENTM discovery engine uniquely enables us to rationally design and develop our diverse library of MGDs and to deploy them against target proteins identified through our QuEENTM discovery engine. Uniquely, many of these target proteins are considered inadequately drugged or completely undruggable by other therapeutic modalities. We actually consider a target protein’s lack of druggability as one of our key criteria for our discovery and development selection and prioritization process. Our resulting MGDs are designed to reprogram the E3 ligase to bind to and induce the degradation of a therapeutically relevant target protein. Central to our QuEENTM discovery engine is a detailed understanding of the molecular interactions promoted by our MGDs between E3 ligases and structural features on the surface of therapeutically relevant proteins, which we refer to as degrons.

Key components of our QuEENTM discovery engine are:

•

AI/ML engines: Our focus on protein surface characterization sets us apart, enabling us to identify reprogrammable E3 ligases as well as potential target proteins amenable to our approach. We have developed sophisticated and proprietary AI-powered algorithms to mine databases of protein sequences and structures, including structures determined from x-ray crystallography and cryoEM, and structures from predicted protein folding. Our proprietary geometric deep learning engine for surface characterization continuously learns from our expanding MGD library, identifying new degrons and surface features (“glueprints”) in targetable proteins across the proteome.

•

High throughput screening, structural biology and proteomics capabilities: We have developed a suite of high-throughput assays that rapidly assess our proprietary MGD library and MGDs generated during specific programs. Coupled with customized automation and robotic systems, our assays can measure ternary complex formation in both a biochemical and cellular format, as well as measure degradation of target proteins in cells, which we use to screen, identify and rapidly optimize our MGDs.

•

Proprietary MGD library: We have built a wholly-owned, proprietary, diverse, and continuously growing chemical library of currently over 75,000 MGDs that we have rationally designed based on our growing expertise in molecular glue anatomy and design, our large proteomics and screening databases, and AI/ML algorithms. Library compounds currently represent more than 1000 unique low molecular weight scaffolds with favorable binding affinities for an E3 ubiquitin ligase.

By capturing our insights and experience with the identification of target proteins amenable to our approach as well as the discovery and development of MGDs through QuEENTM, we are constantly increasing the power of our discovery engine.

Our QuEENTM discovery engine continues to generate discovery stage programs targeting therapeutically relevant proteins otherwise considered undruggable or inadequately drugged. We are progressing our discovery stage programs for multiple other undisclosed target proteins. Our focus is on target proteins that have been considered undruggable or insufficiently drugged, that are highly credentialed preclinically or clinically, and that can potentially move quickly into clinical development in indications with high unmet need and substantial commercial potential.

In October 2023, our wholly-owned subsidiary, Monte Rosa Therapeutics AG, or Monte Rosa AG, entered into a strategic collaboration and licensing agreement with F. Hoffmann-La Roche Ltd., or Roche Basel, and Hoffmann-La Roche Inc., or Roche US, and together with Roche Basel referred herein as Roche. Pursuant to the License Agreement, the parties will seek to identify and develop MGDs against cancer or neurological disease targets using our proprietary drug discovery platform for an initial set of targets in oncology and neuroscience selected by Roche, with each target being subject for a limited time to certain substitution rights owned by Roche. We will lead preclinical discovery and research activities until a defined point. Upon such point, Roche gains the right to exclusively pursue further preclinical and clinical development activities. Under the terms of the agreement, Monte Rosa received an upfront payment of $50 million, and is eligible to receive future preclinical, clinical, commercial and sales milestone payments that could exceed $2 billion, including up to $172 million for achieving preclinical milestones. We are also eligible to receive tiered percent royalties ranging from high-single-digit to low- teens on any products that are commercialized by Roche as a result of the collaboration.

In September 2025, we and Novartis entered into a collaboration, option, and license agreement, under which Monte Rosa granted to Novartis an exclusive, royalty-bearing, sublicensable and transferable license to degraders for one immunology and inflammation, or I&I program, or the First Licensed Program, and the exclusive option to obtain exclusive, royalty-bearing, sublicensable and transferable licenses with respect to two programs from the Company’s growing preclinical immunology portfolio, or the Options, and the programs, or the Optioned I&I Programs. Such Options are individually exercisable at Novartis’ discretion until a program meets criteria for investigational new drug application-filing-readiness. On a program-by-program basis, if Novartis does not exercise an Option, all rights with respect to such program are retained by the Company; if Novartis does exercise its Option, such program becomes a Licensed Program, or together, with the First Licensed Program, the Licensed Programs. Under the 2025 Novartis Agreement, the Company will apply its proprietary AI/ML-enabled QuEENTM product engine for the discovery and development of degraders for the First Licensed Program and the Optioned I&I Programs. The Licensed Programs will be further developed and commercialized by Novartis, unless otherwise agreed to by the parties in accordance with the 2025 Novartis Agreement. Research activities for the Licensed Programs governed by the Agreement will be overseen by a Joint Research Committee.

Under the agreement, the Company received a $120.0 million non-refundable upfront payment from Novartis. The Company is entitled to receive further payments from Novartis to maintain the Options totaling up to $60.0 million, and is also eligible to receive from Novartis (1) preclinical milestone payments relating to the First Licensed Program and option exercise payments related to the Options of up to $180.0 million, (2) up to $5.4 billion in clinical development, regulatory, and sales milestones relating to the First Licensed Program and the two Optioned I&I Programs, beginning upon initiation of Phase 1 studies, including (a) potential development and regulatory milestone payments of up to $2.2 billion if regulatory approval is achieved for multiple indications in multiple territories and (b) potential sales milestone payments of up to $3.2 billion, allocated across licensed products, and (3) tiered royalties on global net sales in the high-single to low double-digit range for the First Licensed Program and in the low double-digit range for the two Optioned I&I Programs. The Company will be responsible for costs related to research activities, while Novartis will be responsible for costs related to development and commercialization activities.

We are led by an experienced team of drug discovery and development experts with deep experience in targeted protein degradation, molecular glues, chemistry, structural biology, data science, disease biology, translational medicine, and clinical development.

Monte Rosa Therapeutics AG, a Swiss operating company, was incorporated under the laws of Switzerland in April 2018. Monte Rosa Therapeutics, Inc. was incorporated in the State of Delaware in November 2019. The Company is headquartered in Boston, Massachusetts with research operations in both Boston and Basel, Switzerland. Our principal executive office is located at 321 Harrison Avenue, Suite 900, Boston, MA 02118 and our telephone number is (617) 949-2643. Information about us is available on our corporate websites at www.monterosatx.com. Information available on our website is not a part of, and is not incorporated into, this Annual Report. We trade on the Nasdaq Global Select Market under the ticker symbol “GLUE”.

Our product pipeline

We have leveraged our QuEENTM discovery engine to generate our pipeline of product candidates with the potential to treat a diverse range of diseases through targeted protein degradation. Our current programs are focused on delivering therapies to target proteins that have been considered undruggable or inadequately drugged in well-characterized biological pathways across clinical indications in immunology, inflammation, cardiovascular diseases, oncology, and other diseases with high unmet needs. We currently retain exclusive

worldwide rights to the programs shown in Figure 1 below, except for MGDs directed against VAV1 including MRT-6160, which we licensed to Novartis in October 2024, and the targets included in the Roche and Novartis relationships.

Figure 1: Monte Rosa Pipeline

Over the next several years, we plan to expand our early-stage product portfolio into our current therapeutic areas of focus and additional therapeutic areas, leveraging the ability of our QuEENTM discovery engine to degrade therapeutically relevant proteins in areas including immunology & inflammation, cardiovascular, metabolic, and genetic diseases.

Our strategy

Our mission is to discover and develop a portfolio of novel small molecule MGDs that selectively eliminate therapeutically relevant proteins. We believe our MGDs have the potential to benefit patients in a broad range of indications with significant unmet medical need. We believe the product candidates identified through our proprietary QuEENTM discovery engine can provide distinct advantages over other modalities, including the ability to address target proteins that have been considered undruggable or inadequately drugged. We intend to fully develop certain programs internally, while also utilizing collaborations to advance programs in areas where we believe that external expertise and financial resources may enable us to more fully realize the therapeutic and commercial potential of a program.

MGDs provide for therapeutic opportunities not constrained by some of the key limitations of conventional small molecule inhibitor drugs. MGDs provide an opportunity to target the vast universe of target proteins without a defined binding pocket, in a highly selective way, due to the diversity of surfaces that can be targeted, resulting in reversible elimination of a target protein. More specifically, because MGDs work by inducing protein-protein interactions between target proteins and an E3 ligase, they do not require a defined binding pocket on the target protein of interest. Thus, MGDs offer a unique opportunity to unlock significant target space and enable us to address target proteins that have been considered undruggable or inadequately drugged. The interaction surfaces we utilize are often not conserved within protein classes and families, allowing us to potentially achieve significant selectivity for our MGD product candidates that we believe is superior to classical small molecule inhibitor drugs. Lastly, we focus on target proteins where experimental evidence suggests that removal of the target is superior to transiently inhibiting it, in particular proteins that have a scaffolding function.

Through our ability to produce potentially highly selective MGDs with fine-tuned speed and depth of degradation, we believe we can generate MGD product candidates with a wide therapeutic window and other therapeutic advantages that may be beneficial in a broad range of indications, including immunology, inflammation, cardiovascular diseases, oncology, metabolic diseases, genetic diseases, and diseases of the central nervous

system (CNS). We believe our platform has the capability to produce MGDs suitable for distribution into any tissue, including MGDs designed to be CNS-penetrant.

In immunology and inflammation, we are uniquely able to target highly credentialed immune signaling proteins in pathologically relevant immune pathways. We prioritize target proteins that are validated through preclinical or clinical (including human genetic) evidence. We have shown that we are able to optimize our MGD product candidates to induce deep and selective degradation of immune-pathway relevant proteins. Our precision oncology programs are focused on the elimination of proteins that are highly validated driver oncogenes in cancer cells (“oncogene addiction”), that define a cancer lineage dependence (“lineage addiction”), or that create a vulnerability specific to tumor cells (“synthetic lethality”).

Key elements of our strategy include:

•

Continue to advance our NEK7-directed MGD, MRT-8102, for the treatment of NLRP3/IL-1/IL-6 driven inflammatory diseases through completion of the GFORCE-1 trial in elevated CVD risk subjects and initiate the GFORCE-2 study of MRT-8102 in ASCVD. MRT-8102 is a potent, highly selective, and orally bioavailable investigational MGD that targets NEK7 for the treatment of inflammatory diseases linked to the NLRP3 inflammasome, IL-1, and IL-6 dysregulation. NEK7 has been shown to be required for NLRP3 inflammasome assembly, activation and IL-1β release both in vitro and in vivo. Aberrant NLRP3 inflammasome activation and the subsequent release of active IL-1β and interleukin-18 (IL-18) has been implicated in multiple inflammatory disorders, including ASCVD, gout, hidradenitis suppurativa, osteoarthritis, asthma, neurodegenerative diseases, and metabolic disorders including metabolic dysfunction-associated steatohepatitis (MASH) and obesity. In January 2026, we reported interim results from our Phase 1 study of MRT-8102. In subjects with increased CVD risk, MRT-8102 demonstrated rapid and durable reductions in systemic inflammation. Specifically, after four weeks of MRT-8102 treatment in subjects with elevated CVD risk and high levels of CRP, CRP levels were reduced by 85%, and 94% of study participants achieved CRP values below 2 mg/L, a threshold associated with reduced CVD risk. We expect to initiate a study of MRT-8102 in in elevated CVD risk patients defined by Stage 3/4 chronic kidney disease and elevated CRP in H2 2026, a Phase 2 study of MRT-8102 in patients with gout flares in Q4 2026 or Q1 2027, and a Phase 2 study of MRT-8102 in patients with moderate to severe hidradenitis suppurativa, in H1 2027.

•

Continue to support Novartis's clinical development of our VAV-directed MGD MRT-6160 in immune-mediated disease. Pursuant to our Agreement with Novartis, Novartis will be responsible for all further clinical development and commercialization of MRT-6160. We believe our global license agreement with Novartis will accelerate and broaden the scope of clinical development of MRT-6160 while retaining substantial value for us, including through milestone payments and our share of the US P&L for MRT-6160 provided under our Agreement;

•

Advance our GSPT1-directed MGD program by initiating a signal-confirming Phase 2 study of MRT-2359 in mCRPC patients with AR mutations. In December 2025, we announced positive interim clinical data from our study of MRT-2359 in combination with enzalutamide in heavily pretreated mCRPC patients, including patients with AR mutations. We presented additional positive data at the ASCO Genitourinary Cancers Symposium in February 2026. In mCRPC patients with AR mutations, treatment with MRT-2359 in combination with enzalutamide led to a 100% PSA response rate in all 5 patients identified as having AR mutations. In addition, MRT-2359 plus enzalutamide demonstrated a 100% disease control rate in this patient subset, per RECIST criteria, including 2 RECIST partial responses and 3 with stable disease. In total, our study included 15 evaluable patients, including the 5 patients identified as having AR mutations. Across those 15 evaluable patients, the overall disease control rate was 67% (10 of 15), with 10 of 15 patients showing tumor size reductions of target lesions, including the 2 RECIST partial response patients that were also identified as having AR mutations. We plan to initiate a Phase 2 study of MRT-2359 in combination with a second-generation AR inhibitor in mCRPC patients with AR mutations, with potential to expand the study into additional patient subsets, including patients naive to 2nd generation AR inhibitors. The study is anticipated to start in 2026;

•

Advance our cell cycle program to IND submission. We believe our programs directed at CCNE1 and CDK2, key drivers of cancers with cyclin dependent kinase pathway alterations, have the potential to achieve greater selectivity for the CCNE/CDK2 complex versus conventional CDK ATP-site inhibitors. We also believe they have the potential to provide more sustained pathway inhibition compared to ATP-site inhibitors. We expect to submit an IND application for a cyclin E1-directed MGD in 2026;

•

Continue to advance and develop our pipeline of rationally designed MGDs to transform the treatment of diseases in multiple therapeutic areas including immunology & inflammation, cardiology, and oncology. Through our QuEENTM discovery engine, we have identified a variety of additional degron-containing proteins that are amenable to our approach and are either undruggable or insufficiently drugged and we continue to build MGDs against these proteins. We continue to advance programs in preclinical development, and to identify new degron-containing target proteins as well as MGDs. We will continue to prioritize therapeutically relevant target proteins backed by strong biological and genetic rationale with the goal of producing novel precision medicines;

•

Continue to enhance and expand the capabilities of our QuEENTMdiscovery engine to unlock the full therapeutic potential of our MGDs in our targeted therapeutic areas. We employ a core set of drug discovery and development principles to guide our target protein selection across various protein classes and therapeutic areas. We are specifically focused on delivering therapies to target proteins that have been considered undruggable or inadequately drugged, and that are situated in preclinically and clinically well-characterized and validated biological pathways;

•

Expand and protect our proprietary know-how and intellectual property.We continue to innovatively expand our intellectual property around our innovations in the field of targeted protein degradation and in particular MGDs. Our intellectual property, which includes proprietary know-how, patent applications and issued and expected patents, as well as trade secrets, applies not only to our product candidates, but also to all of our various innovations, including, for example, our drug discovery processes including our QuEENTM discovery engine; our AI-based E3 ligase characterization algorithms, AI-based degron discovery algorithms, AI-based novel MGD design algorithms, and in silico screening algorithms; our drug development tools; our growing library of MGDs; the innovative methods and approaches we have developed to rationally design MGDs to expand our library, and to certain biomarkers and therapeutic applications for our potential product candidates;

•

Execute our discovery collaboration with Roche in the areas of cancer and neurology. Under the terms of the agreement, Monte Rosa will lead discovery and preclinical activities against multiple select cancer and neurological disease targets to a defined point. Upon such point, Roche gains the right to exclusively pursue further preclinical and clinical development of the compounds. We believe this collaboration will enable and accelerate expansion of our platform into neuroscience and additional areas of oncology; and

•

Execute our collaboration with Novartis for degraders to treat immune-mediated diseases. Under the terms of the agreement, Monte Rosa’s scientists will apply our proprietary AI/ML-enabled QuEENTM product engine for the discovery and development of degraders to be further developed and commercialized by Novartis. Monte Rosa’s publicly disclosed pipeline programs are outside the scope of this agreement.

Background on targeted protein degradation and molecular glue degraders

Proteins drive nearly all biochemical reactions in the body and many diseases stem from abnormal intracellular protein activity. Proteins, including those inside the cell and on its surface, are attractive therapeutic targets; nevertheless, despite advances in therapeutic modalities, approximately 75% of human proteins remain undruggable by traditional small molecule inhibitors.

Challenges with druggable vs. undruggable proteins

Traditional small molecule inhibitors target proteins by binding to a pocket on the protein’s surface. The absence of a binding pocket presents a challenge to the discovery and development of traditional small molecule inhibitors. Indeed, many proteins, including key disease-driving proteins such as transcription factors, scaffolding proteins, and enzyme modulators, often lack druggable pockets, making them undruggable by conventional small molecule inhibitor approaches. Other therapeutic modalities that can target such proteins, such as therapeutic antibodies, oligonucleotide-based therapies, and genetic therapies, are limited in their ability to address aberrant protein behavior. Although these therapies have improved patient outcomes, they face challenges in delivery, scalability, and therapeutic application. A summary of characteristics of various therapeutic modalities compared to MGDs is shown in Figure 2.

Figure 2: Characteristics of Therapeutic Modalities, Including MGDs, the Next Generation of Precision Medicine-Based Small Molecule Drugs

Molecular glues: our expanding approach to protein degradation

Protein degradation is one of the body’s natural processes by which proteins are eliminated from human cells through the attachment of a molecular tag, called ubiquitin, to a protein by any of the approximately 600 human E3 ligases, marking the protein for degradation by the proteasome in the cell. Targeted protein degradation can be mediated by two small molecule classes: MGDs and PROTACs (proteolysis-targeting chimeras, also known as heterobifunctional degraders) (illustrated in Figure 4).

We believe our targeted protein degradation approach offers many features that make it an attractive therapeutic modality:

•

Removal of a target protein: partial or complete removal of a target protein can lead to more complete inhibition of signaling and metabolic pathways, thus resulting in more profound and longer lasting pharmacodynamic effects than traditional reversible or irreversible inhibition can induce.

•

Targeting intracellular proteins: small molecule-based protein degraders, in particular MGDs, readily cross cell membranes or can be optimized to do so.

•

Ease of delivery: small molecule-based protein degraders, in particular MGDs, can be delivered through various routes of administration, including orally.

•

Systemic and tissue distribution: since most small molecule-based degraders, in particular MGDs, are low molecular weight compared to other therapeutic modalities, tissue distribution, such as into the CNS or tumor tissues, poses less of an issue.

•

Catalytic mode of action: after inducing degradation of a target protein molecule, the small molecule-based protein degrader-E3 ligase complex is able to induce the degradation of additional target protein molecules. Thus, the small molecule-based protein degrader acts catalytically, unlike protein inhibition, causing the removal of many target protein molecules with a single MGD molecule, thereby editing the cellular proteome.

•

Event driven pharmacology: unlike with inhibitors where prolonged engagement of the drug with the protein is required for efficacy, small molecule-based protein degraders only require engagement with the E3 ligase and the target protein long enough to induce tagging for degradation.

As described above, there are multiple potential advantages of the protein degradation approach, but one of the most intriguing is the potential to achieve greater therapeutic benefits resulting from the durable but reversible removal of a target protein from the cellular proteome.

Figure 3: Molecular Glue Degraders; Expanding Target Space, Fostering a New Generation of Drugs

Our approach

MGDs are small molecule-based protein degraders designed to modify an E3 ligase’s binding specificity and thus can employ the body’s natural mechanisms of protein destruction to selectively eliminate therapeutically relevant proteins.

Our QuEENTM discovery engine was built for the rational, target-centric discovery of potent and selective MGDs with favorable drug-like properties, thus potentially systematically overcoming common challenges of MGD discovery, as illustrated in Figure 4

Figure 4: QuEENTM is Redefining the Rules of MGD Discovery

We believe our discovery engine has the potential to continue to deliver MGD product candidates, including product candidates that could address target proteins that have been considered undruggable or inadequately drugged, while possessing attractive pharmaceutical properties. As shown in Figure 6, our initial programs utilize

cereblon as the E3 ligase system to tag target proteins. Through our generation of data-at-scale, AI/ML platform and proprietary MGD library we have expanded and continue to expand chemical and target space, and have now begun to leverage other E3 ligase systems.

Figure 5: Our Rational Approach to Unleash the Full Potential of MGDs

We have built our discovery engine on the insight that deep knowledge and understanding of features of protein surfaces drives MGD discovery. Surfaces and their unique features, which we call “glueprints”, mediate protein-protein interactions and targeted protein degradation. As shown in Figure 7, interrogating surfaces using geometric deep learning enables us to identify reprogrammable ligases and the matching target protein space, creating broad potential opportunities to eliminate undruggable, disease-driving proteins through “only-in-class” MGDs.

Figure 6: Surface Interactions Drive "Only-in-Class" MGD Designs

QuEENTM Discovery Engine

We design and develop MGDs in a rational and iterative approach using our industry-leading and dynamic QuEENTM discovery engine, encapsulating our team’s proprietary knowledge and discovery capabilities across biology, chemistry and computational sciences, and from which we are generating our library and pipeline of MGD product candidates. Through our discovery engine, we have built intellectual property that allows us to induce a high degree of surface complementarity between an E3 ligase and a target protein, potentially leading to high potency and selectivity of MGDs for the therapeutically relevant target proteins we select.

The QuEENTM discovery engine was built to support our approach to the discovery and development of MGD product candidates that degrade a wide landscape of therapeutically-relevant target proteins by (i) systematically identifying degrons and other surface features on target proteins that may enable ternary complex formation and consequential degradation initiated by E3 ligases, (ii) understanding how to reprogram the surface of endogenous E3 ligases using small molecule-based MGDs; and (iii) rationally designing MGDs that can be optimized towards high potency and selectivity, with favorable pharmaceutical properties. Our process of degron discovery and MGD design is highly iterative and interdisciplinary. Our quantitative mass-spectrometry-based proteomics and high throughput screening capabilities allow us to screen our library at scale to facilitate our discovery efforts. Powerful AI modeling learns from and guides our high throughput screening and chemo-proteomics, which in turn feed information back to the AI engine, and the accumulated knowledge is used to guide our MGD discovery programs and library expansion. For example, MGD discovery and development for a protein target can pass from degron identification, to MGD hit identification, to in silico improvement, to a round of chemo-proteomics validation, to chemical library alterations and back, until we reach the desired selectivity and degradation.

Our Proprietary MGD library

We discover and develop lead MGDs for degron-containing target proteins by screening our MGD library of currently over 75,000 MGD molecules, and applying proximity screening tools and our chemo-proteomic capabilities in QuEENTM. We continue to expand our highly diverse library of MGDs based on our growing expertise in MGD design, our knowledge of the cereblon-binding surface, and variations in target surface features and degrons. We have developed unique and innovative synthetic chemistry approaches to access over 1,000 scaffolds, each designed to probe three-dimensional structural and chemical property space differently. These scaffolds are being utilized as building blocks to generate our proprietary library of highly diverse compounds. The modular construction of our library allows us to explore different areas of chemical space and follow-up rapidly on hits from our library. Our highly diverse library of MGDs leverages different areas of the cereblon surface to

engage diverse degrons and surface features on target proteins. Our library has given rise to multiple series of MGDs for each of the target proteins currently being studied across our disclosed and undisclosed portfolio.

Our AI/ML engine identifies reprogrammable E3 ligases and E3 ligase-accessible target proteins

Our focus on protein surface characterization sets us apart, enabling us to identify reprogrammable E3 ligases as well as potential target proteins amenable to our approach. We have developed sophisticated and proprietary AI-powered algorithms to mine databases of protein sequences and structures, including structures determined from x-ray crystallography and cryoEM, and structures from predicted protein folding. fAIceit – our proprietary geometric deep learning engine for surface characterization - continuously learns from our expanding MGD library, identifying new degrons and surface features (“glueprints”) in targetable proteins across the proteome.

High throughput screening of our proprietary library identifies active MGDs

We have developed a suite of high-throughput assays that rapidly assess our proprietary MGD library and MGDs generated during specific programs. Coupled with customized automation and robotic systems, our assays can measure ternary complex formation in both a biochemical and cellular format, as well as measure degradation of target proteins in cells, which we use to screen, identify and rapidly optimize our MGDs.

Ourquantitative proteomics profiling assays for neosubstrate identification and MGD optimization

Utilizing mass-spectrometry-based proteomics, we have developed a suite of unbiased high throughput quantitative profiling assays to assess cellular protein degradation, selectivity of degradation, target ubiquitination, and ternary complex formation. Combined with our end-to-end instrument, automation and computational infrastructure, the platform enables us to screen our library, identify new targets amenable to our approach, and drive our drug discovery programs rapidly from hit identification to development candidate.

Ourstructural biology platformenables the rational design of our MGDs

Leveraging high-throughput crystallization and cryo-electron microscopy, we have established a robust pipeline for generating high resolution protein structures. We use structural insights derived from these to support the design of our MGD library, rationally optimize our MGDs during lead optimization programs, and validate novel binding modes target proteins that are highly diverse with regards to the protein-protein interface involved.

Our QuEENTM discovery engine has enabled us to discover novel degrons, protein surface features and binding modes, some of which have been published in scientific journals such as Science, dramatically expanding our addressable target space. We have used our AI engine and a rational design approach to discover MGDs that are exquisitely selective, enabling us to potentially eliminate therapeutically relevant target proteins in pathways that are highly relevant for diseases with high unmet need in immunology, inflammation, cardiovascular diseases and oncology as well as other diseases.

QuEENTM expansion

Our QuEENTM discovery engine was originally focused on identifying and developing MGDs that induce the binding of degron-containing neosubstrates to cereblon as a means of targeting them for degradation. Using our established tools, we are expanding the scope of QuEENTM to further grow the cereblon target space, to leverage additional E3 ligases for targeted protein degradation, and to potentially extend the utility of our degraders to target multiple therapeutic targets.

•

Expand chemical space: We are expanding the diversity and the chemical space covered by our MGD library based on our understanding of protein surfaces. Using structure-based design and AI-driven algorithms we have identified more than 1000 unique scaffolds which form the basis of our MGD library of over 75,000 compounds;

•

Activate new E3 ligases: We believe that we will be able to reprogram other E3 ligases through the discovery of ligase specific MGDs as well as specific ligase-accessible degrons, thus enabling us to generate ternary complexes with a further subset of the approximately 600 E3 ligases;

•

Grow target space: We believe expanding degron identification, identification of other protein surface features, E3 ligase activation, and MGD chemical space will unlock previously undruggable proteins for therapeutic intervention;

•

Explore modality expansion: We believe our experience and capabilities enable us to expand the utility of small-molecule based induced-proximity for unique clinical applications.

Our Approach for Immunologic and Inflammatory Diseases

MRT-6160, a highly selective and orally bioavailable VAV1-directed MGD in development for the treatment of immune-mediated diseases

Overview

VAV1 is a Rho-family guanine nucleotide exchange factor that plays a critical role in T- and B-cell receptor signaling and activity. As many immune-mediated diseases are thought to be driven by an underlying dysregulation or hyperactivation of T- and/or B-cells, a VAV1-directed MGD, which we believe will ameliorate aberrant responses from both cell types, has broad potential application for immune-mediated diseases.

There are multiple published studies providing preclinical data supporting VAV1’s potential as an attractive target for attenuating T- and B-cell activity, as shown in Figure 7. Studies report that VAV1 knockout mice are viable and fertile, but display various loss-of-function T- and B-cell phenotypes, and are protected from experimentally induced autoimmune diseases. In addition, it was shown using whole-genome CRISPR screens in primary human T cells that VAV1 plays a key role in T-cell function and that genetic loss of VAV1 confers loss of IL-2 secretion, amongst other functional consequences.

We believe our VAV1-directed MGDs have the potential to modulate both T- and B-cell function as well as the cross talk between these cell types when activated in autoimmune disease. Despite being a preclinically validated target for attenuating T- and B-cell activity, VAV1 has remained undruggable to date using small molecule inhibitor approaches due to the lack of an appropriate binding pocket for small molecule inhibitor design. Therefore, targeting VAV1 with a VAV1-directed MGD and eliminating its activity through protein degradation could provide therapeutic benefits in multiple T- and/or B-cell mediated autoimmune diseases.

Figure 7: VAV1 is a Highly Validated Target for Attenuating T-cell and B-cell Activity

In October 2024, we and Novartis entered into a License Agreement under which we granted to Novartis an exclusive license to develop, manufacture, and commercialize VAV1-directed MGDs including MRT-6160. We were responsible for completing the Phase 1 clinical study and Novartis is responsible for all subsequent development and commercial activities starting at Phase 2. We received from Novartis an upfront payment of $150 million and are eligible to receive up to $2.1 billion in development, regulatory, and sales milestones, beginning upon initiation of Phase 2 studies and including potential development and regulatory milestone payments, exceeding $1.5 billion if multiple indications achieve regulatory approval in multiple territories. We and Novartis also agreed to a net profit and loss sharing arrangement, in which we will co-fund any global clinical development from Phase 3 onwards and will share 30% of any profits and losses associated with the manufacturing and commercialization of the licensed products in the United States. We are eligible to receive from Novartis potential sales milestone payments in connection with sales outside of the United States, and tiered royalties on sales outside of the United States. Novartis will be responsible for costs associated with any

subsequent clinical studies except for the Phase 3 cost covered by us under the profit and loss sharing agreement.

We have demonstrated, in vivo, that once daily oral dosing of MRT-6160 inhibited disease progression in well-established models of multiple sclerosis, rheumatoid arthritis, inflammatory bowel disease, and spontaneous autoimmune disease such as systemic lupuserythematosus and Sjogren’s disease, as shown in the Figures and discussion below. We believe the public literature, coupled with our data package, summarized herein, provides strong support for use of a VAV1-directed MGD in a broad range of systemic and CNS autoimmune diseases. Based on this support, we advanced our VAV1 development candidate, MRT-6160 into clinical studies. In August 2024, we announced initiation of our MRT-6160 Phase 1 single ascending dose/multiple ascending dose (SAD/MAD) study. Results from the Phase 1 study are provided herein.

Development of VAV1-directed MGDs

A summary of the VAV1 intracellular signaling pathway is illustrated in Figure 8.

Figure 8: VAV1 is a Key Regulator of T- and B-cell Receptor Activity

VAV1 is an upstream signaling node associated with multiple clinically validated pathways impacting immune cell functions, as shown in Figure 9. These include T cell activation, B cell activation and plasma cell differentiation, Th17 response, and pro-inflammatory cytokine production. Therapies targeting these pathways individually have been approved for multiple autoimmune and inflammatory diseases.

Figure 9: VAV1 is an Upstream Targeting Node Associated with Clinically Validated Pathways

MRT-6160 is a first-in-class molecular glue degrader of VAV1. MRT-6160 forms a strong ternary complex with VAV1 and cereblon through a newly characterized non-canonical degron which was unveiled through application of our QuEENTM discovery engine technologies. The unique character of the VAV1 degron and its interaction with cereblon induced by MRT-6160 result in a high degree of selectivity over commonly degraded neosubstrates and other closely related VAV family proteins. Our studies show that MRT-6160 degrades human VAV1 with a DC50 of 7 nM and Dmax of 97%, is orally bioavailable across species, and displays favorable in vitro ADMET properties. The favorable drug-like profile of MRT-6160 is summarized in Figure 10.

Figure 10: MRT-6160 is a Potent, Selective VAV1 MGD Development Candidate with a Favorable Drug-Like Profile

Degradation of VAV1 in peripheral immune cells was observed following MRT-6160 oral administration. Additionally, MRT-6160 has brain penetrance with anticipated dose dependent degradation of VAV1 in the CNS. Non-clinical safety profiling showed a clean profile with respect to mutagenicity (mini-Ames), hERG activity, CYP inhibition and induction, and broad off-target screening (CEREP panel).

Preclinical 28-day GLP toxicology studies in rats and non-human primates (cynomolgus macaque or cyno) demonstrated a highly favorable profile. The no-observed-adverse-effect-level (NOAEL) was set at the highest doses tested in both species. The exposure at NOAEL for rats was approximately 1000-fold over the projected human efficacious exposure, and the exposure at NOAEL for cynos was approximately 600-fold over the projected human efficacious exposure. In healthy cynos, no adverse immunotoxicity or impact on peripheral immune compartments was observed. There was no observed impact on bone marrow and peripheral hematopoietic cell counts. No gastrointestinal toxicity was observed. Furthermore, there were no off-target effects identified in in-vitro safety profiling, no genotoxicity, no phototoxicity, and no hERG activity.

The potency and selectivity profile of MRT-6160 was evaluated in primary human peripheral mononuclear blood cells (hPBMCs). As shown in Figure 11, left panel, MRT-6160 elicited dose-dependent degradation of VAV1 in primary human T and B cell subsets. As shown in Figure 11, right panel, tandem mass tag (TMT)-global proteomics assessment revealed selective degradation of VAV1 over its closely related family members VAV2 and VAV3 in addition to other proteins expressed in hPBMCs and detectable in the assay.

Figure 11: MRT-6160 Selectively Degraded VAV1 in Primary Human Immune Cells

MRT-6160 was further characterized for anticipated on-target pharmacodynamic and functional activity in primary human T and B cells. As shown in Figure 12, in primary human T cells (top panel), VAV1 degradation by MRT-6160 resulted in inhibition of TCR-mediated pharmacodynamic (CD69) and functional activity (IL-2 secretion and proliferation). In primary human B cells (bottom panel), VAV1 degradation by MRT-6160 resulted in inhibition of BCR-mediated pharmacodynamic (CD69) and functional activity (IL-6 and soluble IgG secretion) demonstrating expected on-target activity in disease-relevant cell types.

Figure 12: VAV1 degradation by MRT-6160 Resulted in Inhibition of T- and B-cell Receptor Signaling and Activity

In vivo validation of VAV1 MGD MRT-6160

MRT-6160 was evaluated in various well-established T- as well as T- and B-cell mediated in vivo models of autoimmune disease. In a T-cell-mediated experimental autoimmune encephalomyelitis (EAE) model of multiple sclerosis (Figure 13, left panel), daily oral dosing of MRT-6160 following disease onset inhibited disease progression in a dose-dependent manner comparable to that of supratherapeutic doses of dexamethasone, a corticosteroid used broadly in autoimmune disease. After 6 days of dosing, samples from mice were assessed by western blot for murine (m) VAV1 levels in diseased tissue. Shown in the right panel of Figure 13, MRT-6160 induced dose-dependent degradation of mVAV1 commensurate with inhibition of disease progression.

Figure 13: MRT-6160 Elicited Dose-Dependent Activity in a T-cell mediated Multiple Sclerosis Autoimmune Disease Model

MRT-6160 was also evaluated in a T- and B-cell mediated collagen-induced arthritis (CIA) model of rheumatoid arthritis. Mice were orally administered MRT-6160 daily following disease onset and scored for clinical signs of disease. As shown in Figure 14, left panel, 1 mg/kg MRT-6160 inhibited disease progression comparably to 10 mg/kg anti-TNF-Alpha antibody. The right panel of Figure 14 shows that treatment with MRT-6160 reduced the

serum levels of anti-collagen II IgG1 and total anti-collagen II IgG antibodies, demonstrating inhibition of auto-antibody production.

Figure 14: MRT-6160 Inhibited Disease Progression and Auto-Antibody Production in the Collagen-Induced Arthritis Disease Model

MRT-6160 was also evaluated in a T-cell transfer-induced model of colitis, as shown in Figure 15. In a prophylactic model shown in the left panel, mice were orally administered vehicle or MRT-6160 daily following T-cell transfer. Anti-TNF-Alpha antibody was administered intraperitoneally every third day as a standard of care control. Oral dosing with 1 mg/kg MRT-6160 demonstrated superior disease inhibition compared to 10 mg/kg anti-TNF-Alpha antibody. In a therapeutic model shown in the right panel, mice were orally administered MRT-6160 starting on Day 17 following disease induction. MRT-6160 was compared to two commonly used oral therapies for rheumatoid arthritis, a JAK inhibitor and a S1PR antagonist, as well as vehicle control. MRT-6160 was superior in controlling clinical signs of disease as compared to both active comparators.

Figure 15: MRT-6160 Ameliorated T Cell Transfer-Induced Colitis Equal to or Better than Standard of Care

Figure 16, left panel, shows reduction of inflammation-mediated damage and swelling of the colon following MRT-6160 treatment in the prophylactic T-cell transfer-induced model of colitis. The right panel of Figure 16 shows mesenteric lymph node and colon CD4+ T cell assessment by flow cytometry where MRT-6160 reduced the

frequency of IL-17A+, TNF-Alpha+, and IL-6+ CD4+ T cells, known drivers of inflammatory bowel disease in humans.

Figure 16: MRT-6160 Inhibited Inflammation-Mediated Damage and Cytokine Production in a Model of Inflammatory Bowel Disease

In a preclinical autoimmune disease model characterized by chronic inflammation, autoantibody production, and multi-organ involvement (Figure 17) administration of MRT-6160 resulted in broad activity across an array of disease markers, including attenuated autoantibody levels and reduced skin and kidney pathology. MRT-6160 was equivalent or superior to prednisone or anti-CD40L monoclonal antibody treatments across multiple metrics of disease pathology.

We believe these findings reinforce the breadth of MRT-6160's potential across multiple immune-mediated diseases, including systemic lupus erythematosus, Sjögren’s disease, rheumatoid arthritis, and others.

Figure 17: MRT-6160 Inhibited Disease Progression, Autoantibody Production, and Nephritis in the MRL-Faslpr Lymphoproliferative Autoimmune Model

In summary, these data further highlight the potential of MGDs to potently degrade otherwise ‘undruggable’ proteins such as VAV1, providing an opportunity to treat immune-mediated diseases with a novel, orally dosed modality capable of blocking multiple pathogenic immune and cytokine receptor pathways in parallel, thereby potentially providing greater clinical benefit, a strategy we are planning to continue to pursue through our portfolio of MGDs.

MRT-6160 Phase 1 Study

In a Phase 1 study of healthy volunteers, MRT-6160 was dosed in five SAD dose level cohorts and three MAD dose level cohorts, as shown in Figure 18. All cohorts were randomized and placebo controlled, and over 70 subjects were enrolled in total. The primary endpoint of the study was safety and tolerability of MRT-6160. The secondary endpoints were pharmacokinetic and pharmacodynamic assessments using various readouts in multiple different analytes.

Figure 18: MRT-6160 Phase 1 Healthy Volunteers Study Design and Objectives

VAV1 degradation was assessed by flow cytometry of CD3+ T cells and CD19+ B cells, as shown in Figure 19. In addition, ex vivo activation of whole blood was performed to assess T and B cell functions, including CD69 upregulation on T and B cells measured by flow cytometry, and cytokine secretion measured by immunoassay.

Figure 19: In-Vitro Assay Validation of PD and Immune Cell Functional Testing

Analysis of plasma concentrations of MRT-6160 over time demonstrated a dose dependent human pharmacokinetic profile, as shown in Figure 20. MAD dosing resulted in an approximately two-fold increase in exposure at steady state. No food effect was observed.

Figure 20: MRT-6160 Displayed a Dose-Dependent Human Pharmacokinetic Profile

As shown in Figure 21, MRT-6160 achieved degradation exceeding 90% at all but DL1 of the SAD cohorts and at all MAD cohorts, based on analysis of peripheral blood T cells. Reduction of VAV1 protein levels was sustained, with dose-dependent recovery following cessation of treatment. Similar results were observed in peripheral blood B cells.

Figure 21: MRT-6160 Achieved Dose-Dependent Degradation >90% in Peripheral Blood T cells After Single and Multiple Dose Administration

VAV1 degradation by MRT-6160 resulted in significant functional inhibition of T and B cells following ex vivo activation of T and B cell receptors in cells derived from whole blood, as shown in Figure 22 for all SAD cohorts. MRT-6160 treatment significantly attenuated CD69 upregulation (a marker of immune cell activation) on T and B cells following TCR stimulation, reflecting functional inhibition of both cell types. In addition, MRT-6160 treatment significantly inhibited IL-2, IFN-γ and IL-17A secretion from whole blood derived T cells following ex-vivo activation of T cell receptor signaling, demonstrating reductions of up to 99% from pre-dose levels. MRT-6160 also attenuated IL-6 production by 60-90% across dose levels, and over 80% at all but the lowest dose level, following B cell activation. Alignment with the pharmacodynamic studies above suggests robust functional effects on cytokine production can be achieved with 80% and higher degradation of VAV1.

Figure 22: VAV1 Degradation by MRT-6160 Resulted in Significant Functional Inhibition of T and B Cells

Suppression of CD69 upregulation following single or multiple doses of MRT-6160 and subsequent ex vivo TCR stimulation of whole blood derived cells was significant (>90%) as well as sustained during post treatment observation periods, as shown in Figure 23 (data for selected SAD and MAD dose shown as example). Similar results were observed in peripheral blood B cells following BCR-stimulation.

Figure 23: MRT-6160 Resulted in Sustained Suppression of TCR-mediated CD69 Activation following Single or Multiple Doses of MRT-6160

MRT-6160 demonstrated a sustained effect on TCR-mediated cytokine production following single and multiple dose administration and ex vivo stimulation of whole blood derived cells, as shown in Figure 24. MRT-6160 treatment resulted in significant and sustained suppression of IL-2, IL-17A and IFN-γ secretion from whole blood derived T cells following ex-vivo activation of T cell receptor signaling (data for selected SAD and MAD dose shown as example).

Figure 24: MRT-6160 Resulted in Sustained Suppression of TCR-mediated Cytokine Production following Single or Multiple Doses of MRT-6160

MRT-6160 was well tolerated with no serious adverse events, or SAE, observed. Observed treatment-emergent adverse events, or TEAEs, were mild (82%) or moderate (18%) and self-limiting. Overall TEAE frequency was similar between MRT-6160 and placebo. TEAEs observed in 2 or more subjects treated with MRT-6160 were: in the SAD cohorts, pain from vessel puncture (2); in the MAD cohorts, cough (2), diarrhea (3), feeling hot (4), headache (5), nasal congestion (2), oropharyngeal pain (3) and pyrexia (2).

In summary, the pharmacodynamic and functional ex-vivo studies suggest significant effects on cytokine production can be achieved following treatment with MRT-6160. Furthermore, we believe the levels of VAV1

degradation observed clinically are consistent with levels of degradation required to induce efficacy in the preclinical models tested so far. The functional impact on cytokine production is also consistent with levels predicted to be required to achieve efficacy in humans, based on benchmark clinical data from other compounds.

In summary, we believe the Phase 1 data described here as well as our chronic toxicology package support a clear path into Phase 2 studies and broad potential applications of MRT-6160 in multiple immune-mediated diseases.

NEK7-directed MGDs for the treatment of inflammatory disease

Overview

The NLRP3 inflammasome is a multi-protein complex that functions as a central signaling hub integrating stimuli derived from pathogens, cellular damage, metabolites and metabolic stress, ultimately triggering the production of pro-inflammatory cytokines, and is implicated in numerous inflammatory diseases. Activation of the NLRP3 inflammasome critically depends on NIMA-related kinase 7, or NEK7, a serine/threonine kinase that is essential to trigger the assembly of the active NLRP3 inflammasome complex in a NEK7 kinase-independent manner. As depicted in Figure 25, aberrant NEK7-dependent activation of the NLRP3 inflammasome leads to the release of highly inflammatory mediators, including the cytokines IL-1α, IL-1β, and IL-18, through a form of cell death known as pyroptosis. The NLRP3 inflammasome and above mentioned cytokines have been implicated in multiple inflammation-driven diseases, including ASCVD, gout, hidradenitis suppurativa, pericarditis, osteoarthritis, and obesity. Given the central role of NEK7 in pathological NLRP3 inflammasome activation and in disease initiation and progression, targeted degradation of NEK7 with a highly selective MGD, such as our product candidate MRT-8102, may effectively suppress NLRP3 inflammasome activity at the most upstream intervention point, thus leading to the broadest downstream inhibition of inflammatory signaling possible, thereby potentially inducing more effective disease resolution than currently approved agents.

Figure 25: NEK7 Enables NLRP3 Inflammasome Assembly and Activation, Pyroptotic Cell Death and Release of Highly Inflammatory Cytokines and DAMPs

Extensive clinical data support the relevance of IL-1 and NLRP3 inflammasome signaling across multiple diseases in large therapeutic areas spanning cardiovascular, rheumatologic, respiratory, dermatologic, neurologic, and metabolic conditions. Cytokine-targeting agents that act downstream of the NLRP3 inflammasome, such as rilonacept (an IL-1α/β blocker) and canakinumab (an IL-1β blocker), have demonstrated clinical activity across several of these inflammatory diseases, including recurrent pericarditis and ASCVD, respectively, as illustrated in Figure 26. Despite the reported activity of IL-1 targeting agents like rilonacept and canakinumab, we believe that upstream therapeutic degradation of NEK7 will drive more complete suppression of the full spectrum of inflammasome-driven signals and thus, may offer even greater benefit to patients.

Figure 26: The NLRP3 Inflammasome and IL-1 Signaling are a Clinically Validated Pathway for Inflammatory Diseases

Below, we describe our internal in vitro and in vivo studies supporting the essential role of NEK7 in cytokine production downstream of NLRP3 inflammasome activation across multiple species, including mouse, rabbit, and cynomolgus monkey. We present evidence that MRT-8102 and related NEK7 MGDs can potently and selectively suppress NLRP3 inflammasome activity in disease models, leading to meaningful improvements in disease burden. Finally, we provide data suggesting that existing therapeutics, such as the GLP-1 receptor agonist semaglutide, may partially function through downregulation of NEK7 expression, although sub-optimally, to achieve anti-inflammatory activity and efficacy, further confirming the crucial role NEK7 plays in the pathological activation of the NLRP3 inflammasome.

Identification of a NEK7 degron and NEK7-directed MGDs

The kinase-independent, scaffolding function of NEK7 in activating the NLRP3 inflammasome suggests that degradation of NEK7 could be an effective way to block NLRP3 inflammasome activation. Indeed, we have demonstrated experimentally that removal of NEK7 by MGD-mediated degradation is an efficient way to prevent NLRP3 inflammasome formation and therefore has the potential for deep pathway inhibition through disassembly of the NLRP3 inflammasome.

Our discovery efforts resulted in the identification of MRT-8102 as a first-in-class NEK7 MGD. As shown in Figure 27, MRT-8102 induces a strong ternary complex of NEK7 with cereblon via a canonical G-loop degron (left panel), resulting in profound NEK7 degradation (DC50 10 nM and Dmax 89%; right panel). MRT-8102 is highly selective against known cereblon neosubstrates and more importantly, against other NEK family members (also see below), is orally bioavailable across multiple species tested, and displays favorable in vitro ADMET properties. Non-clinical safety profiling showed a clean profile with respect to mutagenicity (mini-ames), hERG activity, and broad off-target screening (CEREP panel).

Figure 27: MRT-8102 is a Potent and Selective Investigational NEK7 MGD with a Favorable Drug-Like Profile

The amino acid sequence of the NEK7 degron is unique among the NEK family members, indicating the potential to identify MGDs that are highly selective for NEK7. As shown in Figure 28, human peripheral blood mononuclear cells (PBMC) were treated with MRT-8102 for 24 hours, followed by TMT-global proteomic profiling. Profound and highly selective degradation of NEK7 is evidenced by a several-fold decrease in NEK7 protein, without significant changes in other detected proteins. Other NEK family members are highlighted on the volcano plot and were not degraded. Several other cell lines and types, including U937, MM1S, iPSCs, and PBMCs derived from cynomolgus monkeys revealed similarly selective proteomic profiles when treated with MRT-8102.

In a PK/PD study in cynomolgus monkeys, a single oral dose of 10 mg/kg of MRT-8102 was sufficient to achieve deep and sustained NEK7 degradation beyond the PK exposure window of the compound (Figure 28, right panel).

Figure 28: MRT-8102, a Potent and Highly Selective NEK7-directed MGD, Induces Durable Pharmacodynamic Modulation In Vivo

MRT-8102 is differentiated from existing NLRP3-IL-1-IL-6 targeting agents

MRT-8102 is an orally bioavailable molecular glue degrader that is designed to selectively and catalytically degrade NEK7 to suppress NLRP3 inflammasome activity. As compared to small molecule inhibitors of the NLPR3 inflammasome, NEK7 degradation by MRT-8102, as shown schematically in Figure 29, leads to long-lasting inflammasome disassembly and sustained inhibition of cytokine release, most importantly without the on/off pathway inhibition characteristic of inhibitors. Moreover, MRT-8102 is exquisitely selective for NEK7, as confirmed by our proteomics work, which may help minimize the risk of off-target effects.

Figure 29: MRT-8102 Induces Catalytic NEK7 Degradation, Long-lasting Inflammasome Disassembly, and Sustained Inhibition of Cytokine Release

Beyond deeper and more sustained pathway inhibition, we believe targeting NEK7 will provide safety advantages over NLRP3 inhibitors, as NLRP3 has been shown to have inflammasome-independent functions, potentially impacting safety. Toxicities with NLRP3 inhibitors have been reported preclinically and clinically and several NLRP3 inhibitors have been discontinued, potentially due to lack of selectivity and resulting toxicities.

Given that MRT-8102 prevents inflammasome assembly and activation, it is highly effective at suppressing pyroptosis, the pathological event that is ultimately responsible for the release of disease-promoting cytokines, such as IL-1α, IL-1β and IL-18, as well as damage-associated molecular patterns (DAMPs), which are known to be important drivers of the inflammatory process. Although various mono- and bi-specific biologics currently under investigation can robustly target one or more of these cytokines downstream of pyroptosis, ultimately, these classes of therapeutics may be limited by their inability to suppress the full spectrum of disease-relevant cytokines and more importantly the release of DAMPs as highlighted in Figure 30. Furthermore, whereas MRT-8102 selectively reduces the pool of cytokines driven by NLRP3 inflammasome activation, biologics, like IL-1 and IL-6 targeting antibodies may indiscriminately inhibit this pool of cytokines, irrespective of their source, and thus potentially elevate infection risk by impacting immune pathways beyond the NLRP3 inflammasome pathway. Such pathways might include other inflammasomes (e.g. AIM2, NLRC4, NLRP1), protease-mediated IL-1 activation, alternative inflammasome activation downstream of TLR4 and passive release from dying cells.

As shown in Figure 30, right panel, whereas MRT-8102 effectively prevented pyroptosis and DAMP release, anti-IL-1 and IL-6 agents failed to significantly inhibit these processes in stimulated human monocyte-derived macrophages (hMDM), suggesting that mono- and potentially bispecific biologics may incompletely block the multitude of pathological drivers of disease.

Figure 30: MRT-8102 Potently Inhibits Pyroptotic Cell Death in Stimulated hMDM

NEK7 MGDs Demonstrate Compelling Efficacy Across a Number of Disease Models and Species

We showed that sustained NLRP3 inflammasome inhibition through NEK7 MGDs drives compelling improvements in disease severity across multiple models and species. Models tested in the course of our NEK7 program include two independent mouse models of cardiovascular disease (Figure 31, left panel) looking at reductions in cardiac damage, a rabbit-based gout model (Figure 31, middle panel) investigating whether MRT-8102 treatment leads to a decrease in severity of monosodium urate (MSU)-induced flares, and a cynomolgus monkey-based diet-induced obesity model (Figure 31, right panel) investigating the potential of a NEK7 MGD to reduce body weight and liver inflammation as a monotherapy or in combination with semaglutide.

Figure 31: MGD-Mediated NEK7 Degradation Improved Disease Burden Across a Wide Range of Disease Models Spanning Multiple Species

Mouse models of cardiac damage

In an acute myocardial infarction (AMI) study conducted in CRBN-I139V mice (Figure 31, left panel), i.e. mice with a partially humanized CRBN protein sequence, the effects of prophylactic dosing with MRT-8102 were compared with an anti-IL-1R antibody and the NLRP3 inhibitor MCC950. Outcomes were evaluated histologically to assess infarct size. Coronary arterial ligation (60 min), followed by reperfusion led to significantly sized infarcts in vehicle-treated animals at the 24-hour timepoint. In contrast, prophylactic treatment with MRT-8102, an anti-IL-1R antibody and MCC950 significantly reduced infarct size. Collectively, these data support the pathological role of aberrant NLRP3 inflammasome activity in myocardial infarction and the therapeutic value of targeting NEK7 as a differentiated approach.

In a pericarditis model, also established in CRBN-I139V mice, the effects of prophylactic dosing with the same agents evaluated in the AMI model were assessed (Figure 31, middle panel). As neutralization of IL-1 and inhibition of IL-1 signaling is a clinically approved approach to treat recurrent pericarditis, treatment with an anti-IL-1R antibody served as a positive control in this study. Administration of zymosan, a potent immunostimulant, to the pericardium resulted in a significant increase in pericardial effusion relative to sham controls. In contrast, prophylactic treatment with MRT-8102, an anti-IL-1R antibody, or MCC950 significantly reduced pericardial effusion. Notably, MRT-8102 demonstrated superior efficacy compared with anti-IL-1R antibody treatment. Since the anti-IL-1R antibody blocks the action of both IL-1α and IL-1β (as does rilonacept, which is clinically approved for use in recurrent pericarditis), these data suggest that the activity of MRT-8102 in a model of pericarditis is in fact broader than blockade of the downstream IL-1 cytokines alone.

Rabbit rheumatology model of gout

In addition to inflammatory diseases of the heart, MRT-8102 also demonstrated activity in a rabbit model of gout. Following intra-articular injection of MSU crystals, marked joint swelling was observed, peaking at approximately 24 hours. Despite achieving only ~40% degradation of NEK7 in rabbits, due to limited homology between human and rabbit CRBN, prophylactic administration of MRT-8102 on Day -1 resulted in a three-fold reduction in peak joint swelling and accelerated resolution of inflammation, with swelling returning to baseline levels, comparable to rabbits not injected with MSU crystals, by days 3 - 6 (Figure 31, right panel).

Cynomolgus monkey diet-induced obesity model

In addition to acute inflammatory indications, MGD-mediated NEK7 degradation also demonstrated activity in chronic metabolic diseases such as obesity. In a cynomolgus monkey model of diet-induced obesity, animals with body mass index ≥ 40 kg/m2 were selected from a colony of animals maintained on a high-fat diet and randomized across treatment groups. A NEK7 MGD with similar properties to MRT-8102 was used for this study. Following single-agent NEK7 MGD treatment for 11 weeks, an approximately 8% reduction in body weight was observed relative to vehicle. When combined with the GLP-1 receptor agonist semaglutide, >23% body weight loss was achieved relative to vehicle treatment (Figure 32, left panel). In addition to monitoring total body weight, dual-energy X-ray absorptiometry (DEXA) body composition analysis was also performed. Notably, the combination treatment demonstrated preferential activity in central abdominal fat, a region associated with elevated metabolic risk, with proportionally greater reductions observed relative to other fat depots such as the gynoid region (Figure 32, middle, right panels).

Figure 32: NEK7 MGD, Alone or in Combo with Semaglutide, Drives Preferential Abdominal Fat Loss While Sparing Lean Mass in a DIO Cyno Monkey Model

Intriguingly, while, as expected, >90% degradation of NEK7 was observed by Western blotting in PBMCs in the NEK7 MGD treatment arms, a significant reduction in NEK7 levels of about 40 – 50% was also detected in the semaglutide single agent treatment arm (Figure 33, left panel). These findings may suggest a mechanism through which GLP-1R agonists suppress NLRP3 inflammasome activity analogous to NEK7 MGD treatment, albeit less potently and reliably. Consistent with the central role of NEK7 in NLRP3 inflammasome signaling, NEK7 levels across treatment arms correlated with changes in plasma IL-18, a canonical inflammatory cytokine associated with NLRP3 inflammasome activity (Figure 33, right panel).

Figure 33: Anti-inflammatory Activity of NEK7 MGD and Semaglutide May Partially Overlap at the Level of NEK7 Degradation/Downregulation and NLRP3 Inflammasome Inhibition

Collectively, we believe these studies demonstrate that targeted degradation of NEK7 by MRT-8102 or other NEK7-directed MGDs enables deep and sustained suppression of NLRP3 inflammasome activity across multiple

species and disease models, including cardiovascular injury, gout, and metabolic diseases. We believe these studies validate a broad and pivotal role of NEK7 in the pathogenic activation of NLRP3 inflammasomes.

In conclusion, by intervening upstream of where cytokine blocking biologics act, we believe that NEK7 degradation may offer the potential for broader and more durable therapeutic benefit across a wide range of inflammation-driven diseases and potentially carry less risk of infections.

Market Opportunities for NEK7-directed MGDs

ASCVD

We believe there are multiple attractive opportunities for NEK7-directed MGDs across a wide range of NLRP3 inflammasome-driven indications, including a particularly attractive opportunity in ASCVD. While LDL cholesterol lowering agents are a well-established part of the treatment paradigm, patients that achieve their LDL-C targets still experience up to a 40% chance of life-threatening cardiovascular events (Holtrop et al. European Journal of Preventive Cardiology, 2024). We believe this demonstrates the substantial residual risk not fully addressed by LDL-C lowering and speaks to the promise and importance of complementary approaches such as targeting the NEK7/NLRP3 pathway.

The role of the NLRP3 inflammasome and IL-1β in ASCVD has been well established through various approaches, including through the key findings of the landmark CANTOS clinical trial. In this study, canakinumab, a monoclonal antibody targeting IL-1b was dosed in over 10,000 patients with prior myocardial infarction and high sensitivity C-reactive protein (hsCRP, a marker of inflammation) levels of > 2 mg/L, a threshold above which there is higher risk of cardiovascular events and mortality. Treatment led to a significant reduction in hsCRP and a significantly lower rate of recurrent cardiovascular events than in placebo treated patients, independent of lipid lowering. Despite the significant efficacy noted, canakinumab was also associated with a higher incidence of fatal infections than was placebo, which ultimately yielded an unfavorable risk-benefit profile. The higher risk of infections is likely due to the above-mentioned indiscriminate and deep suppression of an NLRP3 inflammasome-independent pool of IL-1b, a cytokine that plays a critical role in host protective immunity elicited by multiple different pathways.

We believe upstream targeting of the NEK7/NLRP3 pathway may have greater potential than downstream IL-6 biologics in ASCVD. Figure 34 shows how monocytes, upon uptake of oxidized LDLs, become the initiator and driver of disease through chronic activation of NLRP3 inflammasomes and consequential pyroptosis. This pathological activation of NLRP3 inflammasomes promotes plaque destabilization and downstream CV events through contribution of cellular debris, lipids and further recruitment of bone marrow derived macrophages to the growing plaques. Given the essential involvement of the NLRP3 inflammasome in this process, we believe that suppression of pyroptosis through degradation of NEK7 and disassembly of the NLRP3 inflammasome with MRT-8102 could be an effective means by which to stabilize plaques and hence reduce cumulative incidence of MACE.

Figure 34: NLRP3 Inflammasome Activation Promotes Plaque Growth, Destabilization and CV Events

As shown in Figure 35, an unbiased analysis using an in-house generated NLRP3 inflammasome activity signature, using our proprietary BreakthruTM data science engine across more than 1000 datasets spanning

hundreds of diseases identified ASCVD as one of the top-ranking conditions with strong NLRP3 inflammasome activation. As expected, cryopyrin-associated periodic syndromes (CAPS), a group of rare and hereditary autoinflammatory disorders, also scored highly in our analysis. CAPS is known to be driven by constitutively active NLRP3 mutants, providing a positive control to this analysis. Interestingly, in addition to CAPS and ASCVD, we noted other diseases, such as gout and hidradenitis suppurativa (HS), also scored highly in our analysis, suggesting MRT-8102 or next generation molecules could offer broad therapeutic value across these indications.

Figure 35: ASCVD, HS and Gout Rank Amongst Top NLRP3 Inflammasome Activated Indications

Consistent with the transcriptomic study, an unbiased genetic association analysis, also performed through our BreakthruTM data science engine and shown in Figure 36, found an NLRP3 gain-of-function single nucleotide polymorphism, or SNP, to be significantly associated with increased downstream CV events including stroke, coronary artery disease, and peripheral artery disease, further supporting the role of the NLRP3 inflammasome in driving disease pathology. By comparison, although IL-6 signaling was also associated with CV outcomes, the effects were weaker than those observed for NLRP3, implying that upstream targeting of NLRP3 inflammasome activity may have the potential for greater efficacy than targeting downstream cytokines such as IL-6. In summary, these data strongly support an opportunity for MRT-8102 in the treatment of CV indications and the management of elevated CVD risk.

Figure 36: Human Genetics Supports Causal Relationship Between NLRP3 and ASCVD

Gout

Gout is a painful, chronic inflammatory arthritis driven by elevated uric acid levels that result in MSU crystal deposition in joints. These crystals potently activate NLRP3-inflammasome-mediated inflammation, leading to recurrent flares with intense pain. Nearly 30% of gout patients are comorbid with stage 3/4 chronic kidney disease (CKD), a condition that both elevates uric acid levels and increases gout risk. Current treatment options present significant limitations in this population. Many therapies are contraindicated in CKD or lack long-term safety data, often requiring dose titration and close monitoring. The anti-IL-1β antibody canakinumab, approved for treatment

of gout patients including those with CKD comorbidity, has demonstrated efficacy in resolving acute flares, further validating the role of the NLRP3 inflammasome in the disease pathology of Gout. However, due to safety concerns, such as the risk of serious infections, canakinumab is not approved for prophylactic use.

Conversely, the urate-lowering therapy pegloticase is approved for chronic, refractory gout as a prophylactic treatment. Yet nearly 70% of patients experience flares within the first three months of therapy. Together, these examples highlight the persistent treatment challenges in the prophylactic setting, coupled with safety concerns in CKD patients, underscoring the need for new and safer therapeutic approaches capable of resolving acute flares while also preventing recurrent flares.

Our studies demonstrated that MRT-8102 significantly reduced MSU crystal–induced caspase-1 activation and downstream pyroptosis, with greater potency than the NLRP3 inhibitor selnoflast, as shown in Figure 37. By blocking pyroptosis, MRT-8102 may not only attenuate inflammatory severity but also preserve macrophage viability, potentially enhancing phagocytic clearance of MSU crystals, the underlying trigger of disease pathology. Consistent with these in vitro findings, MRT-8102 demonstrated robust activity in a rabbit model of gout as previously described in Figure 31 (middle panel). In addition to improving joint swelling, MRT-8102 treatment also led to statistically significant reductions in pathologic musculoskeletal ultrasound findings and histopathology scores at the end of the study, further supporting a robust resolution of gout flares (Figure 36; bottom right).

Collectively, these data, together with the clinical validation of IL-1 pathway inhibition through canakinumab, support a compelling potential development opportunity for MRT-8102 as a differentiated therapeutic for gout—either as a single agent or in combination with standard-of-care urate-lowering therapies.

Figure 37: MRT-8102 Inhibits NEK7/NLRP3 Pathway and Has Potential to Resolve and Prevent Gout Flares

Hidradenitis suppurativa (HS) Hidradenitis suppurativa (HS) is a chronic, complex inflammatory skin disease characterized by painful, recurrent nodules and abscesses that may progress to sinus tract formation and fistulae. Molecular profiling of HS lesions has identified several signaling pathways that contribute to disease pathogenesis, leading to the approval of targeted therapies such as anti-TNF agents (adalimumab) and anti-IL-17 therapies (secukinumab and bimekizumab). Despite these advances, physicians estimate that approximately 55% of patients remain inadequately controlled on current standards of care, suggesting that additional inflammatory pathways contribute meaningfully to disease pathology.

HS is strongly associated with metabolic syndrome, including dyslipidemia, obesity, and insulin resistance, highlighting a potential role for metabolically driven inflammation in disease pathogenesis. In this context, the NLRP3 inflammasome is thought to be an important mediator of disease pathology, as it can be activated by metabolic danger signals such as elevated free fatty acids and hyperglycaemia. Consistent with published findings, our internal data science analyses demonstrate that NLRP3 inflammasome activity is markedly elevated in HS lesions, at levels comparable to or exceeding those observed in atherosclerotic plaques (Figure 35).

Notably, NLRP3 inflammasome activity remains significantly higher in anti-TNF non-responders compared with responders, suggesting that NLRP3-mediated inflammation may represent one of the dominant pathogenic drivers in this subset of patients (Figure 38). Further supporting this concept, the dual anti-IL-1α/β antibody lutikizumab recently demonstrated positive Phase 2 results in moderate-to-severe HS, particularly among patients who had previously failed anti-TNF therapy. Treatment with 300 mg every other week resulted in significantly higher HiSCR75 response rates (≥75% reduction in abscesses and inflammatory nodules) compared with placebo, along with meaningful reductions in skin pain at Week 16.

Figure 38: NLRP3-IL-1b Axis is Significantly Active in HS Lesions, Particularly in the Anti-TNF Non-responder Setting

Collectively, these data support a pathological role for aberrant NLRP3 inflammasome activation in HS and highlight a compelling development opportunity for MRT-8102 in both anti-TNF–naïve and anti-TNF–refractory settings, either as monotherapy or in combination with currently approved agents. Moreover, given the efficacy observed for NEK7 MGDs in diet-induced obesity models (Figure 32), MRT-8102 may reduce upstream metabolic danger signals, further reinforcing NLRP3 inhibition and potentially enabling more durable disease control.

MRT-8102 toxicology studies suggest considerable safety margin

In 28-day repeat-dose GLP toxicology studies in male and female rats and cynomolgus monkeys, no MRT-8102 related clinical signs, nor changes in immunophenotyping, and no gross or clinical pathology findings were observed at any dose level. Therefore, the no-observed-adverse-effect levels (NOAEL) were established at the highest doses tested in these studies, respectively. Additional IND-enabling GLP safety studies did not indicate significant safety concerns related to in vitro off-targets, mutagenicity, phototoxicity, hERG, or in vivo respiratory or CNS safety pharmacology (assessed in rats) and cardiovascular safety pharmacology (assessed in cynomolgus monkeys).

In a long-term toxicology study in cynomolgus monkeys, deep and sustained pathway inhibition was well tolerated following daily dosing of MRT-8102 for three months. There were no test-article related findings, and the NOAEL was determined as the highest dose tested. No body weight loss, unscheduled deaths, or clinically meaningful changes in hematology or clinical chemistry parameters were observed. Furthermore, no gross pathological findings were identified throughout the duration of the study.

In totality, our preclinical safety assessments suggest an at least 200 to 300 fold therapeutic margin over the projected human dose.

MRT-8102 Phase 1 SAD and MAD Study Interim Clinical Results

Based on the promising preclinical profile of MRT-8102, in July 2025, we initiated dosing in a Phase 1 combined SAD and MAD study (Figure 39). The primary endpoint was safety and tolerability. Key secondary and exploratory endpoints included pharmacokinetics and inflammatory markers, including assessment of NEK7

degradation in peripheral blood T cells by flow cytometry, changes in the acute-phase reactants hsCRP and fibrinogen, levels of endogenous IL-6 in blood and cerebrospinal fluid, and IL-1β levels following ex vivo stimulation.

Interim results were based on a data cut-off date of December 23, 2025. In the SAD cohorts, we enrolled 48 participants across 5 dose levels ranging from 40 mg to 400 mg, and in the MAD cohorts, we enrolled 40 participants across 5 dose levels ranging from 5 mg to 200 mg. The 40 mg dose cohort of the ongoing Part 3 of the study, which evaluates MRT-8102’s activity at 40 mg for 28 days in subjects with elevated CVD risk, is expected to enroll approximately 36 subjects. As of the data cutoff date, 24 subjects had completed 4 weeks of dosing and CRP assessment.

Figure 39: MRT-8102 Phase I Study – Dose Levels and Endpoints

Consistent with our preclinical studies, we observed rapid and marked degradation of NEK7 in peripheral blood T cells following a single administration of MRT-8102 (Figure 40). Using flow cytometry, approximately 80-90% NEK7 degradation was noted at 6 hours post a single dose of MRT-8102, a level that was sustained following multiple administrations, ranging from 7 days in the MAD portion of the trial to up to 4 weeks in the Part 3 portion of the study. In the MAD portion, a dose as low as 5 mg achieved considerable degradation of NEK7 24h after the last of 7 doses, and levels of degradation at 5 mg where similar to those at the higher doses tested. These results were consistent with preclinical data from our cyno PK/PD studies, which also suggested that at these levels of NEK7 degradation deep pathway inhibition, including suppression of IL1b upregulation on ex vivo stimulation, can be achieved.

Figure 40: MRT-8102 Achieved 80 – 90% NEK7 Degradation in Peripheral Blood T Cells After Single and Multiple Dose Administration

In light of achieving NEK7 degradation of 80 to 90% across all dose levels, we analyzed downstream inflammatory markers, including hsCRP, in aggregate across all SAD dose level cohorts. In this analysis, we observed a significant reduction in high-sensitivity CRP after a single dose of MRT-8102, as shown in Figure 41. Across all subjects treated with a single dose of MRT-8102, most of whom had normal CRP levels at baseline, we observed a 52% reduction in CRP at 96 hours post-dose. As expected, the reduction in CRP was greater in subjects with higher median baseline CRP, with median CRP reductions at 96 hours of 72% and 78% in the subsets with baseline CRP ≥ 1 mg/L and ≥ 2 mg/L, respectively. Comparable activity was noted across all SAD dose levels, ranging from 40 to 400 mg, suggesting the potential for a wide range of doses to be available for future development.

Figure 41: Single Dose of MRT-8102 Led to Significant Reduction in Serum hsCRP

Comparable results were obtained from the MAD (7 days of dosing) portion of the study, detailed in Figure 42. Similar to the SAD part of the study, we were able to analyze all MAD dose level cohorts in aggregate, based on comparable NEK7 degradation levels achieved across 5 to 200 mg. Through this analysis, we observed a 61% reduction in CRP in all subjects treated with MRT-8102, and greater reductions in subjects with elevated baseline median CRP levels, approaching nearly 80% in subjects with median baseline CRP of ≥1, ≥2, or ≥3 mg/L. Of note, individuals with CRP levels of 2 mg/L or higher are at greater risk of CV events; lowering CRP levels below this threshold is crucial for reducing CV morbidity and mortality.

Figure 42: Multiple Daily Doses of MRT-8102 Led to Significant and Sustained Reduction of Serum hsCRP

Based on 7 days of treatment, 7 of 9 (78%) subjects with baseline CRP ≥ 2 mg/L achieved hsCRP suppression to < 2 mg/L, indicative of the potential to lower CV risk, as shown in Figure 42. The frequency of subjects’ hsCRP value dropping to < 2 mg/L was 67% and 78% depending on whether a baseline level of >= 3 mg/L or >= 2 mg/L was used as a cut off (Figure 43).

Figure 43: Multiple Daily Doses of MRT-8102 led to significant proportion of subjects achieved hsCRP reduction to <2 mg/L*

To gain a deeper, mechanistic understanding of MRT-8102’s impact on CRP, we monitored plasma levels of the pro-inflammatory cytokine IL-6, a well-characterized stimulator of CRP production and secretion from the liver, as detailed in Figure 44. Consistent with the previously noted reduction in CRP, MRT-8102 treatment significantly reduced IL-6 by 55% in the 14 MAD subjects with a median baseline CRP ≥ 1 mg/L. Importantly, the absolute levels of IL-6 were reduced below the threshold of 1.65 pg/mL defined by the previously mentioned CANTOS study, a level below which a significant decrease in the risk of CV events and mortality was reported.

Knowing that the NLRP3/IL-1b axis stimulates IL-6 production, we also investigated the impact of MRT-8102 on IL-1β production and secretion in ex vivo whole-blood stimulation experiments. Consistent with the strong reduction in CRP observed in subjects with elevated baseline CRP, and despite relatively modest induction levels under the assay conditions used, we observed a comparable, near 80% inhibition of IL-1β secretion in whole-blood ex vivo assays from these subjects following multiple administrations of MRT-8102. Importantly, we observed a correlation between NEK7 degradation and IL-1β levels throughout the dosing period, as shown here for a representative subject that displayed about 80% degradation and close to 90% inhibition of IL-1β (80.6% and 87%, respectively) at day 5 post dosing. In summary, MRT-8102, during the MAD portion of our Phase 1 study,

effectively inhibited the entire NLRP3-IL-1β-IL-6-CRP axis, reducing critical biomarkers to levels associated with a significantly reduced risk of CV events and mortality.

Figures 44: Multiple Daily Doses of MRT-8102 Led to Reductions of IL-6 and IL-1β

We sought to characterize the impact of MRT-8102 on CSF inflammatory markers, including CRP and IL-6, following administration of MRT-8102 at a single dose level of 100mg (Figure 45). CSF levels of MRT-8102 were consistent with levels needed to be active against NEK7 (data not shown). Importantly, in two subjects with elevated IL-6 in cerebrospinal fluid at baseline, MRT-8102 administration resulted in reduced CSF IL-6 levels after 7 days of dosing. Plasma IL-6 levels at baseline in these two subjects were low, suggesting a centrally (CNS)-driven mechanism for the elevation as well as the suppression of CSF IL-6 levels.

Figure 45: MRT-8102 Treatment Reduced IL-6 Levels in CSF Consistent with CNS Penetration

Part 3 of the study (CRP PoC cohort) evaluated a cohort of subjects with elevated CVD risk. The study design, shown in Figure 46, included 36 subjects randomized 3:1 to MRT-8102 at a dose of 40mg once daily or placebo. Subjects were defined as having elevated CVD risk based on measures of obesity and elevated plasma CRP levels. The primary endpoints were safety and tolerability, with secondary endpoints including changes in CRP levels and pharmacokinetics. Endpoints for hsCRP included absolute reduction and frequency of reduction to < 2 mg/L, a threshold that defines lower CVD risk, as discussed earlier. We also measured pharmacodynamic markers, including NEK7 degradation (see above), as well as levels of IL-6, IL-18, and fibrinogen in plasma.

Figure 46: CRP PoC (Part 3) Study of MRT-8102 in Subjects with Elevated CVD Risk

Data analysis included data from 24 subjects (including both MRT-8102 and placebo) who had completed 4 weeks of dosing as of the data cutoff date. As shown in Figure 47, the interim data suggests that MRT-8102, dosed at 40 mg once daily, induced rapid and deep reductions of hsCRP and fibrinogen. The panel on the left shows that median CRP declined by 80% after one week, consistent with our observation from the MAD portion of the Phase 1 study, and by 85% after four weeks of dosing. Also, 94% of subjects reached hsCRP levels below 2 mg/L, meaning their hsCRP values returned to levels associated with lower CVD risk, as shown in the middle panel. Lastly, there was a 31% reduction in fibrinogen, an independent atherosclerotic risk factor, observed during the treatment period.

Figure 47: Analysis of CRP PoC Cohort Suggested MRT-8102 Induced Rapid Reductions of hsCRP and Fibrinogen

As of the data cut-off of December 23, 2025, 112 subjects had completed dosing across the SAD, MAD, and Part 3 portions of the study. Across this sample of patients, blinded review showed a favorable safety profile with no SAEs. Treatment-emergent AEs were mild to moderate. There was no evidence of increased infection risk or dose-dependent AEs. The evaluation and data collection are ongoing for Part 3 of the study. Of note, one participant in Part 3 was diagnosed with asymptomatic, acute infectious hepatitis A while on study. Because the data were blinded, it was not known whether the participant received MRT-8102 or placebo. The participant experienced a transient ALT elevation equivalent to a Gr 3 that improved while continuing on treatment for several days.

Review of unblinded data from 88 participants from the SAD/MAD cohorts confirmed that single and multiple doses of MRT-8102 were safe and well tolerated up to and including the highest dose level. No SAEs were noted,

and no TEAEs were over grade 2. The treatment arm reported TEAEs in 29% of participants and the placebo arm reported TEAEs in 32% of participants. The most frequent TEAE was headache, which was reported in 9% of participants in both the placebo and treatment arms.

Based on these highly encouraging initial data, the MRT-8102 Phase 1 study, now named GFORCE-1, was expanded to include additional dose exploration of MRT-8102 in subjects with elevated CVD risk (Figure 48). We are enrolling subjects in 3-dose cohorts, randomized 3:1 to active drug vs. placebo, for a total of approximately 108 subjects. The ongoing cohort at 40 mg will be one of the 3 dose levels. Based on the interim clinical data suggesting that effects on CRP levels can be induced early and were sustained over the course of 4 weeks, we believe our 28-day study can provide valuable information on the dose levels necessary to achieve therapeutic benefit and guide our subsequent GFORCE-2 Phase 2 study in ASCVD, which we anticipate will initiate in 2026. Lastly, we believe our expanded GFORCE-1 study will also provide key insights for development of MRT-8102 and other potential NEK7 MGD product candidates in additional indications.

Figure 48: GFORCE-1 Study: Dose Exploration of MRT-8102 in Subjects with Elevated CVD Risk

We believe MRT-8102 has strong therapeutic potential in a broad range of inflammation-driven diseases and we are planning for broad development of our NEK7 MGDs, including MRT-8102 and potentially other NEK7 MGD product candidates, across this space. We plan to prioritize development of MRT-8102 in ASCVD and then to expand development of MRT-8102 or a next generation NEK7 MGD product candidate into other indications that meet our internal criteria, including unmet medical need as well as self-developability.

We are planning our next study of MRT-8102 in elevated CVD risk patients defined by Stage 3/4 chronic kidney disease and elevated CRP, as shown in Figure 48, and we have named this study GFORCE-2. This is a proposed trial design subject to review by FDA. GFORCE-2 is expected to initiate in H2 2026 and will evaluate the effect of MRT-8102 treatment for up to 12 weeks (followed by open label extension of 12 weeks) on CRP levels, as well as impact on liver fat, liver inflammation, and obesity, as illustrated in Figure 49. GFORCE-2 aims to generate additional safety data with longer dosing of MRT-8102, and to confirm the effects on CRP levels seen in data from GFORCE-1. Also, we believe the trial has the potential to generate data useful for understanding MRT-8102’s potential in other, related indications such as MASH and obesity.

Figure 49: GFORCE-2: Phase 2 Study in Elevated CVD Risk Patients Defined by Stage 3/4 Chronic Kidney Disease and Elevated CRP

We are also planning to initiate a Phase 2 study of MRT-8102 in patients with acute gout flares, as shown in Figure 50. This is a proposed trial design subject to review by FDA. We anticipate this study to initiate in either Q4 2026 or Q1 2027. The study is expected to randomize approximately 40 patients with recurrent single joint gout flares to 12 weeks of treatment with one of two doses of MRT-8102. Outcome measures include reduction of pain Visual Analogue Scale (VAS) by 72 hours and frequency of new flares.

Figure 50: Phase 2 Study in Acute Gout Flares

We expect to initiate a Phase 2 study of MRT-8102 in patients with moderate to severe hidradenitis suppurativa in H1 2027. Outcome measures are expected to includeHiSCR75 after 16 weeks of MRT-8102 treatment relative to placebo. This is a proposed trial design subject to review by FDA.

Our Precision Medicine Approach for Cancer

MRT-2359, a highly selective and orally bioavailable GSPT1-directed MGD in development for the treatment of AR and MYC-driven Prostate Cancer

Overview

GSPT1 (also known as eRF3a) is a translation termination factor that catalyzes protein synthesis termination, facilitating the release of mRNA and newly synthesized protein from the ribosomal machinery. We have identified GSPT1 as a potential therapeutic vulnerability in MYC-driven cancers with high protein translation activity, including androgen receptor (AR)-positive prostate cancer.

MRT-2359 is an orally bioavailable MGD that, through extensive in vitro and in vivo studies, we have shown to induce the degradation of GSPT1. MRT-2359 is designed to preferentially affect the growth and survival of MYC family transcription factor-driven cancer cells addicted to protein translation. In vivo, once-daily oral dosing of MRT-2359 led to potent antitumor activity in MYC-driven cell-line-derived and patient-derived xenograft models. MRT-2359 is currently in a Phase 1/2 clinical trial (ClinicalTrials.gov Identifier: NCT05546268). Based on our

preclinical and clinical work to date, we are developing MRT-2359 in metastatic castration resistant prostate cancer (mCRPC).

Development of GSPT1-directed MGDs to Target Downstream Vulnerabilities of MYC Activation

It is well established that abnormal activation of MYC, for example through translocation or high expression, results in uncontrolled cell growth, associated with increased protein synthesis and a ramp-up of the protein translation machinery. MYC-driven tumors are therefore widely believed to be addicted to protein translation, which creates an inherent dependence on critical components of the translation machinery, such as GSPT1, as illustrated in Figure 51. As part of our research program, we identified GSPT1 as a potential novel vulnerability in MYC-driven cancers, demonstrating that degradation of GSPT1 leads to inhibition of the MYC pathway and downregulation of the expression of critical oncogenic signaling molecules and pathways, in particular in AR-positive prostate cancer cells.

Figure 51: Targeting MYC-driven Tumors and Their Addiction to Protein Translation Through GSPT1 Degradation

Targeting GSPT1 with MRT-2359 (preclinical data)

MRT-2359 is a potent and selective GSPT1-directed MGD discovered and rationally designed using our QuEENTMdiscovery engine. Key features and various pharmaceutical parameters of MRT-2359 are shown in Figure 52.

Figure 52: MRT-2359 is a Selective and Orally Bioavailable GSPT1-directed MGD Rationally Designed Using our QuEENTM Discovery Engine

As shown in Figure 53, MRT-2359 displays preferential activity in MYC-high cancer cells by optimally reducing protein translation through degradation of GSPT1 by 60-70% (left panel, top right). In contrast, higher levels of GSPT1 degradation, as achieved with MRT-2136, lead to a more pan-toxic behavior with narrowed preferential activity in MYC high versus MYC low expressing cells (left panel, bottom right).

Figure 53: MRT-2359 has Optimized Depth of Degradation to Achieve Preferential Activity in MYC High Cancer Cells

Prostate Cancer as an Attractive Target Indication for MRT-2359

Prostate cancer is the second leading cause of cancer-related death among men. The androgen receptor (AR) signaling axis, widely recognized as an important driver of prostate cancer growth, has long been targeted through castration and other systemic therapies. Although androgen deprivation therapy (ADT) is initially effective, resistance almost invariably develops, leading to a more aggressive disease state known as castration-resistant

prostate cancer (CRPC). A defining feature of metastatic CRPC (mCRPC) is the persistent activation of AR signaling through mechanisms such as AR gene amplification or overexpression, the emergence of constitutively active AR splice variants, and AR mutations. While second-generation AR-directed therapies, including enzalutamide, have improved overall survival and radiographic progression-free survival in mCRPC in both pre- and post-chemotherapy settings, responses remain limited in patients harboring AR alterations, underscoring the need for novel therapeutic strategies for this patient population.

Beyond AR signaling, multiple oncogenic pathways are known to contribute to prostate cancer progression and survival. Notably, MYC plays a central role in several aspects of prostate cancer biology (Figure 54), including activation of E2F and AR transcriptional programs, and has been implicated in resistance to multiple therapeutic modalities, including AR inhibitors and radioligand therapies. Given the complex and redundant oncogenic landscape of prostate tumors, therapeutic approaches capable of simultaneously targeting multiple oncogenic nodes may enhance treatment responses and help overcome mechanisms of therapy resistance.

Early in our discovery and preclinical development efforts, we identified mCRPC as a promising indication for MRT-2359. MRT-2359 demonstrated robust inhibition of tumor cell growth and viability in prostate cancer cell lines, including both those sensitive and resistant to anti-androgen therapies. Mechanistically, MRT-2359 reduced the expression of MYC and other key oncogenic proteins through GSPT1 degradation and inhibition of translation. We therefore hypothesized that rational combination strategies incorporating MRT-2359 with second-generation AR inhibitors or other approved agents, such as radioligand therapies, may enhance therapeutic efficacy and maintain activity against CRPC tumors, including those harboring AR mutations or other alterations.

Figure 54: MRT-2359 Exploits Key Therapeutic Vulnerabilities in Therapy-Resistant CRPC

Supporting our mechanistic hypothesis is data evaluating hundreds of cancer cell lines spanning multiple tumor lineages, demonstrating that prostate cancer lines co-expressing high levels of MYC and AR present much greater sensitivity to MRT-2359 than prostate cancer cells that are low or negative for MYC or AR, or neuroendocrine prostate cancer cell lines (Figure 55). Furthermore, in addition to baseline expression of the above mentioned oncoproteins, MRT-2359 treatment also led to significant down-modulation of MYC, AR and E2F signaling selectively in AR and MYC high expressing cell lines but not in PC3 cells (AR negative and MYC low), as assessed by RNAseq.

Figure 55: Pharmacogenomic Profiling Identified AR/MYC-positive Prostate Cancer Cell Lines as Exquisitely Sensitive to MRT-2359

In support of the above-mentioned transcriptomic analysis that showed significant down-modulation of multiple oncogenic pathways selectively in MRT-2359 responsive cell lines, an unbiased global proteomics analysis across the responsive lines revealed significantly reduced cellular abundance of AR (wild type [WT] and the genetic variant AR-V7), MYC and Cyclin D1 proteins, potentially shedding light on the biological mechanism through which MRT-2359 treatment reduces signaling output through these oncoproteins and their downstream pathways (Figure 56). Notably, MRT-2359, as a single agent, demonstrated deeper reductions in AR activity than the AR antagonist enzalutamide in the cell lines tested in vitro within the 24 hour timeframe of the experiment, suggesting MRT-2359 activity toward this critical lineage-defining pathway may be achieved more rapidly than with the currently approved AR-directed therapeutic.

Figure 56: Proteomics Analysis Revealed Modulation of AR and MYC/E2F Pathway by MRT-2359

Consistent with pharmacogenomic screening and in vitro validation studies, MRT-2359 demonstrated encouraging single-agent as well as combination activity with enzalutamide across a number of prostate cancer cell line-derived xenograft (CDX) models, as shown in Figure 57. These models include LNCaP, a cell line characterized by high-level expression of the homozygous T878A mutation in AR, and VCaP, a cell line with amplification of AR WT and low-level expression of the constitutively active V7 isoform. In both models, suboptimal dosing of MRT-2359, when combined with enzalutamide, drove significant tumor regressions and substantially outperformed enzalutamide single agent treatment. Since MYC can promote expression of DNA repair genes, potentially blunting response to radioligand therapy, we also tested the potential for MRT-2359 to improve activity of a PSMA-based radioligand therapy, as shown in the right graph. Whereas single-agent

treatments achieved stasis at best, combination treatment led to significant regressions, suggesting strong synergy between MRT-2359 and the PSMA-based radioligand.

Figure 57: MRT-2359 in Combination with Enzalutamide or Pluvicto Demonstrated Increased Activity Over Monotherapies in Xenograft Models

MRT-2359 Phase 1/2 Clinical Study Design and Preliminary Results

Based on our encouraging preclinical data demonstrating that MRT-2359 combined well with both AR inhibitors and radioligand therapy to improve preclinical activity across CDX models spanning multiple AR alterations, we assessed the potential of MRT-2359 to benefit patients with metastatic CRPC post multiple lines of prior treatment.

Figure 58 outlines the design of our Phase 1/2 clinical study. We conducted robust dose exploration in the monotherapy arms to confirm multiple safe dose levels and to determine our initial recommended Phase 2 starting dose of 0.5 mg daily on a 21-day on-drug, 7-day off-drug schedule.

Figure 58: MRT-2359 Phase 1/2 Clinical Study Design

We initiated the Phase 2 expansion cohort in heavily pretreated, metastatic CRPC patients and reported results for 23 patients that had been enrolled as of the January 30 2026 data cut-off. Notably, we required RECIST measurable disease for all patients entering the trial, a more stringent requirement than is typical for prostate cancer studies, which results in enrollment of a more severe, and often more heavily pretreated patient population with extensive metastatic disease.

For our metastatic CRPC expansion cohort, we applied a multipronged approach to characterize tumors at the molecular level, including the identification of key AR alterations in all enrolled patients. To do so, we employed RNA and DNA sequencing of tumor biopsies, and molecular testing of circulating tumor DNA and circulating tumor cells to detect AR alterations such as AR mutations and/or splice variants (Figure 59). We also verified adenocarcinoma biology through RNA sequencing, which allowed us to exclude from the efficacy analysis tumors that had transformed to primarily neuroendocrine status.

Figure 59: Biomarker Profiling of Tumor and Liquid Biopsies

Figure 60 details the patient demographics, clinical characteristics and prior therapies of patients in the study. Patients in this study were more heavily pretreated than in comparable studies in mCRPC patients. In our study, 78% of patients had been previously treated with second generation androgen receptor inhibitors, 83% with chemotherapy, and 57% with Pluvicto. Of the 23 patients enrolled as of the data cutoff, 15 were evaluable for efficacy. 1 patient had not yet received an on-treatment scan prior to the data cutoff date, 2 patients were non-evaluable due to early consent withdrawal, 1 patient was not evaluable due to investigator decision, 1 patient was non-evaluable due to early clinical progression, and 3 patients were excluded from data analysis based on molecular profiling of their baseline biopsies showing transformation to neuroendocrine differentiation.

Figure 60: Patient Demographics, Clinical Characteristics, and Prior Therapies

Figure 61 highlights the safety and tolerability profile of MRT-2359 in combination with enzalutamide, as of the data cut-off date. The combination was generally well tolerated, and the safety profile observed was favorable when our data was compared with third-party data on other drugs emerging as combination agents for metastatic CRPC, including EZH2 inhibitors. One patient had a dose-limiting toxicity (DLT) (grade 3 stomatitis associated with pain). The most common treatment-related AEs for MRT-2359 plus enzalutamide were fatigue (N=12, 52%), diarrhea (N=11, 48%), and nausea (N=8, 35%) which were classified as mild or moderate and were manageable and not therapy limiting. No dose discontinuations were observed due to AEs.

Figure 61: Treatment-related Adverse Events Occurring in >20% of Patients

Figure 62 shows the waterfall plot for PSA responses, a frequently used marker of therapeutic benefit in prostate cancer, as of the data cut-off date. Patients were categorized by their AR status (WT, gray; V7, blue; mutant, purple). Of note, among the 15 evaluable patients, 5 had AR mutations, and all 5 achieved a PSA response. This includes 2 PSA90 responses and 3 PSA50 responses.

Figure 62: Best Change in PSA in AR-wild type, AR-V7 and AR-mutant Patients

Figure 63 shows the RECIST waterfall plot for the 15 evaluable patients as of the data cut-off date. Two of the 5 AR mutant patients achieved a RECIST partial response (PR), one confirmed and one unconfirmed, and the other 3 presented stable disease (SD), resulting in a disease control rate (DCR) of 100% in this patient subset. Of all 15 evaluable patients, the overall RECIST DCR was 67% (10 of 15), with 10 of 15 patients presenting tumor size reductions of target lesions versus baseline scans, including all 5 patients whose tumors were harboring AR mutations (including the 2 patients with a RECIST PR).

Figure 63: Best Change in RECIST in AR-wild type, AR-V7 and AR-mutant Patients

As shown in the swimmer plot in Figure 64 displaying months on treatment as of the data cut-off date, treatment effects in these heavily pretreated patients were durable in several patients. In the AR mutant subset, 2 patients remained on therapy for 10 cycles or longer and 2 of 5 patients remained on drug as of the data cutoff on January 30, 2026.

Figure 64: Swimmer Plot of All Evaluable Patients

Figure 65 provides additional data on the 5 patients with AR mutations, as off the data cutoff date, where the MRT-2359/enzalutamide combination led to robust and durable PSA and RECIST responses. Looking at these patients in isolation, all patients previously received abiraterone, 3 of 5 previously received a second-generation AR inhibitor, 4 of 5 received the PSMA-targeting radioligand therapy Pluvicto, and 5 of 5 were previously treated with chemotherapy. Consistent with the 100% PSA response rate and 100% disease control rate in this subset of patients, mutant allele frequency in ctDNA and total circulating tumor cell counts were also significantly decreased in 4 of 5 patients, with data unattainable for the fifth patient due to poor sample quality.

Figure 65: Best % Change in PSA, Sum of Diameters of Target Lesions, Variant Allele Frequency and CTC Counts in AR-mutant Patients

To identify signaling pathways associated with tumor size reductions and RECIST responses, we performed an unbiased analysis of pre-treatment biopsies (Figure 66, left panel). Consistent with our therapeutic hypothesis, we found that MYC, E2F, and AR signaling were among the top upregulated pathways associated with the magnitude of tumor size reductions, results that align clinically with what we had seen in preclinical experiments and further support the proposed therapeutic mechanism.

We also assessed whether combination treatment could suppress signaling through these pathways. Consistent with preclinical observations, we observed a significant decrease in output through these oncoproteins and pathways, as shown in Figure 66 (right panel), with, for example, reduced E2F signaling in 5 of 6 on-treatment biopsies.

Figure 66: Analysis of Tumor Biopsies Provides Proof-of-Modulation of Target Oncogenic Pathways

Figure 67 provides a case study of a patient with an AR H875Y mutation, including data as of the data cut-off. This patient had previously been treated with several therapeutics with multiple mechanisms of action, including chemotherapy, radioligand therapy, and an investigational bispecific antibody. Despite the number of prior treatments, MRT-2359, when combined with enzalutamide, led to a RECIST response that correlated with rapid and sustained decreases in blood PSA and in the AR H875Y allele frequency in ctDNA. While PSA values began to rebound at ~cycle 8 of treatment, tumor target lesion size continued to decrease, consistent with the mechanism of action of MRT-2359 described above, which extends to modulation of non-AR pathways such as the MYC and E2F pathways.

Figure 67: Confirmed RECIST PR and PSA90 Response in mCRPC Patient with Activating AR Mutation

In Figure 68, we highlight the treatment journey of a second patient with an AR mutation, including data as of the data cut-off. At baseline, this patient harbored the AR L702H mutation. This patient was also heavily pretreated, having received chemotherapy, enzalutamide, Provenge and several other therapies. Despite these treatments and the advanced stage of disease, the patient responded favorably to the MRT-2359/enzalutamide combination. Blood PSA and mutant allele frequency in ctDNA were significantly decreased after 3 months of treatment, correlating with a decrease in the sum of target lesions. Interestingly, as with the prior patient case study, although PSA began to rebound at cycle 6, tumor regression, as assessed by RECIST, was maintained until cycle 10 of treatment. We believe this supports the conclusion that MRT-2359, at least in part, exerts its activity through an AR-pathway-independent mechanism.

Figure 68: Confirmed RECIST SD and PSA50 Response in mCRPC Patient with Activating AR Mutation

Based on the data we have obtained to date, we plan to conduct a signal-confirming Phase 2 study of MRT-2359 in combination with a second-generation AR inhibitor in patients with AR-mutant tumors. The study, planned to

begin in 2026, will enable further efficient evaluation of MRT-2359 in metastatic CRPC using a Simon two-stage design. Based on the favorable AE profile and clinical activity observed to date in the Phase 2 arm of our Phase 1/2 study, we intend to evaluate a 0.5mg dose of MRT-2359 administered on a 21-day-on, 7-day-off schedule. The study is expected to enroll up to 25 patients with metastatic CRPC harboring AR mutations in their tumors, and there is potential to evaluate additional patient subsets, including patients naïve to 2nd-generation AR inhibitors, if activity in the AR-mutant patient population is confirmed. The primary Phase 2 study endpoints will be PSA response, RECIST response, duration of response, radiographic progression-free survival, and safety. Data from this study could confirm MRT-2359’s clinical activity in relevant patient groups, further clarify its role in the metastatic CRPC treatment landscape, and position the program for advancement into registrational studies.

Cyclin E1-directed MGD molecules for the treatment of cancer

Cyclin-dependent kinase protein complexes (cyclin-CDK) regulate progression through the cell cycle, whereby different combinations of the two subunits control different stages of the cell cycle. They are formed by an association of a regulatory subunit, a cyclin, with an inactive catalytic (kinase) subunit, a cyclin-dependent kinase (CDK). Once the complex is formed, it transitions into an active state, whereby the catalytic or CDK subunits become productive and can phosphorylate downstream effector substrates.

Cyclin E proteins, encoded by the CCNE1 and CCNE2 genes, complex with CDK2 to form an active cyclin E-CDK2 complex, regulating G1-to-S transition of the cell cycle and initiation of DNA replication, as shown in Figure 69. Under normal conditions, cyclin E expression is tightly regulated and restricted to the G1-S phase of the cell cycle. However, many cancer types, including ovarian, endometrial, gastric, and breast cancers, bear frequent amplification or overexpression of the CCNE1 gene, resulting in increased cyclin E1 protein expression and aberrant regulation of cell growth. As such, cyclin E1 represents a genuine oncogenic driver and cyclin E1 amplified cancers are greatly dependent on sustained high levels of cyclin E1 for their continued growth and survival. Hence, pharmacologic suppression of high cyclin E1 protein levels, for example through MGD induced degradation is expected to inhibit tumor growth, in line with the classical “oncogene addiction” paradigm.

Figure 69: CCNE1 (Cyclin E1) Drives Multiple Hallmark Cancer Mechanisms and is a Target for Solid Tumors with Deregulated Cyclin E1

As a regulatory subunit with no catalytic activity, cyclin E1 has been considered “undruggable” to date. We have identified multiple MGD molecules that selectively promote the association of cyclin E1 and cereblon in vitro, while sparing the cyclin E2 paralog. These compounds have shown the ability to robustly and selectively induce cyclin E1 degradation in multiple cancer cell lines in vitro and in disease relevant models in vivo. In addition, they suppress cancer cell line proliferation preferentially when CCNE1 is amplified and/or overexpressed, suggesting robust biomarker-driven activity.

In vitro data

As shown in Figure 70, the cyclin E1-directed MGD MRT-55811 selectively degraded cyclin E1, led to downstream pathway suppression, and induced robust G1/S cell cycle arrest.

Figure 70: MRT-55811 is highly selective and showed biological activity in CCNE1-amplified cell lines

As shown in Figure 71, in addition to concentration-dependent cyclin E1 degradation, MRT-55811 treatment also led to parallel reduction of both phosphorylated and unphosphorylated forms of CDK2, as well as the downstream RB phosphorylation. Mass spectrometry assessment of ubiquitinated peptides following MRT-55811 treatment revealed that both cyclin E1 and the associated CDK2 protein were ubiquitinated, suggesting co-degradation of both components of the cyclin E1-CDK2 holoenzyme.

Figure 71: MRT-55811 Induced CCNE1-CDK2 Holoenzyme Degradation in CCNE1 Amplified Cell Lines

MRT-55811 showed superior differential suppression of tumor growth in CCNE1 dependent cell lines compared to clinical development-stage CDK2 inhibitors and S-phase protein kinase inhibitors (WEE1 inhibitor azernosertib and PKMYT1 inhibitor lunresertib),or clinical stage CDK4/6 inhibitors, as shown in Figure 72. Unlike MRT-55811, several tested clinical stage CDK2, WEE1, or PKMYT1 inhibitors did not fully recapitulate genetic dependency, potentially indicating off-target activity.

Figure 72: MRT-55811 Exhibits Superior Selectivity for Cancers with High CCNE1

In vivo data

When dosed orally as a single agent in preclinical cell line-derived xenograft models of CCNE1-amplified ovarian cancer, gastric cancer, and breast cancer, the cyclin E1-directed MGD MRT-55811 induced robust tumor growth suppression and regression in all three models, as shown in Figure 73.

Figure 73: MRT-55811 Treatment Resulted in Tumor Regression in CCNE1 Amplified Models

We expect to submit an IND application in 2026 for a cyclin E1-directed MGD.

CDK2-directed MGD molecules for the treatment of cancer

Cyclin dependent kinases, or CDKs, are a family of closely related kinases that regulate progression through the cell cycle. CDK activity is modulated by specific cyclins. For example, cyclin E1 binding activates CDK2, as shown in Figure 74. Importantly, increased activity and reliance on CDK2 due to cyclin E1 overexpression is thought to be one of the key mechanisms of resistance occurring in ER+ breast cancer patients when treated with CDK4/6 inhibitors such as ribociclib. Therefore, we believe that selective elimination of CDK2 using CDK2-directed MGDs may provide benefit to these patients. Previously reported small molecule inhibitors and PROTACs of CDK2 have been limited in their selectivity due to the high degree of similarity among the active sites of kinases, in particular within the CDK family itself. We have identified multiple MGD molecules that selectively promote the association of CDK2 and cereblon in vitro, while avoiding other CDKs. Through ongoing lead optimization chemistry, the most advanced compounds are orally bioavailable and can robustly and selectively induce CDK2 protein degradation in multiple cancer cell lines in vitro and in disease relevant models in vivo, leading to strong tumor growth inhibition.

Figure 74: CDK2 is One of the Key Regulators of the Cell Cycle

Lead optimization towards orally bioavailable CDK2-directed MGDs

Our CDK2-directed MGDs form a strong ternary complex with CDK2 and cereblon through a newly characterized non-canonical degron which was unveiled through application of our QuEENTM discovery engine technologies. The unique character of the CDK2 degron interaction with cereblon, and the optimized features of our MGDs provide a high degree of selectivity over closely related proteins such as CDK1, CDK4, and CDK9. Our MGDs are designed to be orally bioavailable with favorable invitro ADMET properties and preclinical safety profiles.

In vitro data

Our lead CDK2-directed MGD MRT-51443 has shown the ability to selectively degrade CDK2 and reduce E2F pathway proteins in vitro, with no significant effect on other CDKs or other kinases, as shown in Figure 75. Our data also support that our CDK2 MGD MRT-51443 can block DNA replication during S phase in CDK2 dependent cells and inhibits cellular proliferation in a concentration-dependent manner.

Figure 75: CDK2-directed MGD MRT-51443 is Selective and Showed Biological Activity in a CDK2 Dependent Cell Line

MRT-51443 displayed superior selectivity compared to clinical CDK2 inhibitors, as shown in Figure 76. Clinical-stage CDK2 inhibitors show off-target activity in biochemical kinome profiling. CDK2 inhibitors, but not a CDK2 MGD, display CDK2-independent activity, as demonstrated by their suppression of cell proliferation in the absence of their primary target, CDK2.

Figure 76: CDK2-directed MGD Displayed Superior Selectivity Compared to CDK2 Inhibitors

Increased activity and reliance on CDK2 due to cyclin E1 overexpression is thought to be one of the key mechanisms of resistance occurring in ER+ breast cancer patients when treated with CDK4/6 inhibitors such as ribociclib. As shown in figure 77, the combination of MRT-51443 and ribociclib delayed resistance onset in in-vitro long term culture assays using a ER+ breast cancer cell line, suggesting that addition of a CDK2 MGD to standard of care therapy might have the potential to delay the occurrence of relapses in patients.

Figure 77: CDK2 MGD/Ribociclib Combination Delayed Resistance Onset

In vivo data

As shown in Figure 78, when dosed orally in preclinical models of ER-positive/HER2-negative breast cancer, MRT-51443 drove deep tumor regression in a triple combination with a CDK4/6 inhibitor (ribociclib) and endocrine therapy (fulvestrant) and substantially reduced tumor burden versus ribociclib + fulvestrant combination therapy alone.

Figure 78: CDK2 MGD Demonstrated Activity in Combination with CDK4/6 Inhibitor and Fulvestrant in ER+ Breast Cancer Model

Other programs

We are specifically focused on developing product candidates for target proteins that have been deemed undruggable or inadequately drugged. Our QuEENTM discovery engine was purpose-built to support the discovery

and development of drugs that degrade a wide landscape of therapeutically relevant proteins by (i) systematically identifying therapeutically relevant target proteins that may be amenable to molecular glue-based degradation; and (ii) rationally designing MGD molecules that can be optimized towards high potency and selectivity, with properties that we believe to be favorable, so to become MGD product candidates. Our pipeline includes programs in I&I indications as well as in oncology. We also have early-stage efforts in areas including cardiovascular, metabolic and genetic diseases.

GLUE US Equity

GLUE · 10-K · period ended 2025-12-31

Item 1. Business

GLUE US Equity☆

GLUE · 10-K · period ended 2025-12-31

Item 1. Business

GLUE US Equity