Scribe Therapeutics ($SCTX) Stock Hub: a CRISPR Epigenetic Silencer Reaches the Clinic With Cash Into 2029

01Latest verified status

Scribe Therapeutics is a clinical-stage company in the narrowest sense of the phrase: one product candidate is in humans, dosing began in mid-2026, and no efficacy result from that trial exists yet. The company describes itself as engineering purpose-built in vivo CRISPR technologies, and its initial focus is the three lipid drivers of atherosclerotic cardiovascular disease: LDL cholesterol, lipoprotein(a) and triglycerides. The lead asset, STX-1150, targets the first of the three by silencing the PCSK9 gene in the liver without cutting or rewriting DNA.

The quarter that closed on June 30, 2026 is a pre-IPO quarter, and reads like one. Cash, cash equivalents and marketable securities were $43.0 million, down from $58.0 million at December 31, 2025. Collaboration revenue was $1.9 million against $4.9 million a year earlier. Research and development expense was $8.8 million against $13.9 million, general and administrative expense $2.5 million against $2.6 million, and the net loss $6.5 million, or $2.62 per basic and diluted share, against $9.9 million and $4.08. The per-share figures are computed on the small pre-IPO share base and are not comparable with anything that follows the July listing.

What changed after the quarter closed. The company effected a 5.9218-for-1 reverse stock split on July 17, 2026, converted its only debt instrument into equity the same day, priced an upsized IPO at $15.00 on July 23, began trading on the Nasdaq Global Market on July 24, and closed the offering and a concurrent Sanofi private placement on July 27 for roughly $155.5 million of aggregate gross proceeds and about $140.7 million net. As of August 31, 2026 the Form 10-Q cover reports 18,931,680 shares outstanding.

Research posture: a technology and financing story that is now funded well past its first human readout, and a scientific question that stays completely open until that readout arrives. The stock trades on a preclinical evidence base and on the credibility of the platform, not on any measured human effect.

02Investment summary and the central debate

The central question with Scribe is not whether the company can pay for its work. After July it can, on its own guidance, into the first half of 2029. The question is whether epigenetic silencing behaves in a human liver the way it behaved in monkeys. Every other consideration, from the size of the PCSK9 opportunity to the value of the Sanofi and Lilly relationships, sits downstream of that single unresolved point.

What makes the question interesting rather than merely open is the shape of the preclinical data. The durability claim is not a projection from a six-week experiment: the lowest dose tested in non-human primates has held LDL-C more than 50% below baseline for two years, as stated in the September 2, 2026 results release; at the May 2026 presentation the figure was more than 22 months and still running. If a fraction of that duration reproduces in humans, the product is not competing with monoclonal antibodies dosed every two to four weeks or with an siRNA dosed twice a year; it is competing with the idea of chronic dosing itself. If it does not reproduce, the company owns an expensive way of doing what inclisiran already does.

03ELXR: what the technology actually does, and what it deliberately does not do

Scribe’s silencing platform is called ELXR, for Epigenetic Long-term X-Repressor. It is built on a nuclease-inactivated version of the company’s engineered CasX protein, fused to epigenetic effector domains. Where a conventional CRISPR editor cuts DNA and lets the cell repair the break, and where base and prime editors rewrite letters of the sequence, ELXR leaves the sequence untouched and instead installs histone and DNA methylation marks at a target promoter. The gene remains present and intact; the cell stops transcribing it.

That design choice is the whole argument. Permanent edits carry permanent consequences, including any off-target edit that happens along the way, which is a difficult trade for a preventive medicine that would be given to broadly healthy people with high cholesterol. A silencing mark, in principle, is durable but reversible, and does not create an irreversible genomic liability. The company frames ELXR as combining the durability of CRISPR with the reversibility of RNA interference, and positions it explicitly against both permanent editing and chronically dosed siRNAs and antisense oligonucleotides.

The engineering claim on top of that is specificity. In an August 31, 2026 bioRxiv publication the company described what it calls an allosterically gated sequential proofreading mechanism, and reported that the engineering approach produced molecules with up to ten-fold fewer off-target effects and as much as four-fold higher on-target activity across loci. The company states that, to its knowledge, this is the first epigenetic therapy with such a mechanism. Preprints are not peer-reviewed, and this one should be read as a description of the company’s own work rather than as independent validation.

04STX-1150 and the PCSK9 thesis

STX-1150 consists of a messenger RNA encoding the engineered ELXR protein and a single guide RNA directed at the PCSK9 promoter, both encapsulated in a liver-tropic lipid nanoparticle licensed from Acuitas Therapeutics. It is given as an intravenous infusion. Once inside hepatocytes, the guide RNA brings the repressor to a specific site in the PCSK9 promoter, transcription of the gene falls, circulating PCSK9 protein falls with it, LDL receptors survive longer on the hepatocyte surface, and more LDL cholesterol is cleared from the blood.

The target itself carries essentially no biological risk, which is the strongest single feature of the programme. PCSK9 is one of the most thoroughly validated targets in cardiovascular medicine: two approved monoclonal antibodies and one approved siRNA already work through it, and human genetics provides the natural experiment. People born with loss-of-function variants in PCSK9 have meaningfully lower lifetime LDL-C and, as the company cites, up to 88% lower risk of coronary heart disease, with no distinguishable adverse effect from a lifetime of low LDL-C. Scribe’s stated ambition is to recreate that genotype pharmacologically, in adults, with one infusion.

The commercial argument rests on adherence rather than potency. The company cites evidence that between 50% and 70% of patients discontinue LDL-lowering medication within a year, and argues that the cumulative nature of arterial injury makes early and uninterrupted lowering worth far more than late or intermittent treatment. A therapy that removes the adherence variable entirely would be worth something distinct from a therapy that simply lowers LDL-C by a large percentage on the day it is measured.

05The preclinical package: what it proves and what it cannot

The core dataset was presented in a late-breaking oral session at the 94th European Atherosclerosis Society Congress, held in Athens from May 24 to 27, 2026, by chief executive Benjamin Oakes, and released publicly on May 28, 2026. In non-human primates, a single administration of an STX-1150 prototype produced circulating PCSK9 reductions of up to approximately 90% and LDL-C reductions of up to 68%. At the lowest tested dose, 0.75 mg/kg, LDL-C fell by more than 50% and, as the September 2, 2026 results release puts it, that reduction has been sustained for two years, with liver enzyme profiles comparable to saline controls. At the May 2026 presentation the same result was described as more than 22 months and still running. A GLP toxicology study in primates showed no test-article-related adverse findings across the dose levels evaluated, and in vitro work found no off-target gene expression changes at three times the EC90.

Two additional details matter for dose translation. The company reports that CMC-scaled STX-1150 was at least five-fold more potent in primary human hepatocytes than in primary cynomolgus hepatocytes, which it presents as supporting a favourable human dose. And in transgenic mice carrying human PCSK9, the effect saturated at doses at or above 0.6 mg/kg. The prospectus states that the expected starting dose in the human trial is 20 mg. It also reports repeat administration at 0.75 mg/kg producing sustained LDL-C reductions of greater than 40% to greater than 60% over 180 days, which is the only preclinical evidence bearing directly on the re-dosing question.

The durability figure is the one that carries the valuation, and it is also the one with the least protection against disappointment. Primate and human liver biology are close but not identical, and the specific worry the company names in its own risk factors is that epigenetic marks may be diluted or erased over time by chromatin remodelling and cell turnover. A single monkey cohort followed for two years is a strong signal from a small and non-human sample. It is not a substitute for a human follow-up curve.

06The Phase 1: design, sites, endpoints and what data will actually look like

Scribe announced clearance from the Australian Therapeutic Goods Administration on May 21, 2026 and initiated the study in mid-2026. The registry record, NCT07428473, lists Monash University as sponsor with Scribe Therapeutics as collaborator, a start date of June 2026, and recruiting status. The principal investigator is Stephen Nicholls, director of the Victorian Heart Institute at Monash University, which places the study with a cardiology group that runs large lipid trials rather than with a generalist phase 1 unit.

Read the endpoint structure carefully, because it defines what the first-half-2027 disclosure can and cannot contain. The primary endpoints are safety endpoints measured to week 52; LDL-C and PCSK9 are secondary. An early data release from the single-ascending-dose portion will therefore be a safety readout with pharmacodynamic evidence attached, from small cohorts, without a control arm and without long follow-up on the later doses. That is normal for the stage. It also means the first data point will show depth of effect and initial tolerability, while the durability question that justifies the platform will only be answerable in the twelve-month follow-up that runs well beyond it.

07Pipeline map: one molecule in humans, two funded to follow

Two points about the shape of this table. First, the wholly-owned programmes are all one therapeutic thesis expressed three ways: lower a genetically defined lipid driver of atherosclerosis, durably, with one dose. That is capital-efficient in that a single delivery system, a single manufacturing approach and a single clinical infrastructure serve all three, and it is concentrated in that a delivery or platform failure would hit everything at once. Second, the prospectus is more precise than the headlines: it anticipates initiating a Phase 1 for one of STX-1200 and STX-1400 as early as 2027 and the other in 2028, not both in 2027.

08The CIRM money: non-dilutive, but not unconditional

On June 18, 2026 Scribe announced awards of more than $25 million from the California Institute for Regenerative Medicine through its Preclinical Development Program. The 10-Q gives the precise structure: aggregate commitments of approximately $25.7 million, split into about $13.0 million for STX-1400 and about $12.7 million for STX-1200, under notices of award entered in May and June 2026, with project periods running from April 2026 to September 2029 for STX-1400 and to March 2029 for STX-1200. The company received $3.8 million of grant cash in the second quarter and carries a $1.8 million grant liability.

The conditions are worth stating plainly, because non-dilutive funding is often described as free money and this is not quite that. Scribe is required to provide aggregate co-funding of approximately $7.1 million from non-CIRM sources, CIRM holds back 5% of payments, and the awards carry future revenue-sharing payments and royalties, described as generally at low single-digit percentage rates. The accounting treatment matters too: the company records the grants as a reduction of research and development expense rather than as revenue, having early adopted the relevant standard. That flatters the reported expense line without flattering revenue, and it should be remembered when comparing research spending across periods.

09Competition: the target is validated, which is exactly why it is crowded

The reason PCSK9 is attractive is the reason it is contested. The approved landscape already includes two monoclonal antibodies, evolocumab and alirocumab, and the siRNA inclisiran, which is dosed twice yearly after loading. Oral PCSK9 inhibitors are in late development at Merck and AstraZeneca. Against that, a one-time therapy has to justify a very different risk profile, and it has to do so in a payer environment that is used to paying for chronic lipid drugs.

Within genetic medicine the prospectus names the direct comparators itself. On gene editing it lists CRISPR Therapeutics, Editas, Intellia, Caribou, Beam, Prime Medicine, Tessera, Mammoth, Arbor and Metagenomi, and notes at least one other gene-editing programme targeting PCSK9 in development at Verve Therapeutics, now a Lilly subsidiary. On epigenetic modification specifically, the named competitors are nChroma Bio, Epicrispr Biotechnologies and Tune Therapeutics, with a set of ex-US cardiometabolic players including AccurEdit, CorrectSequence, Epigenic Therapeutics and Yoltech. In Lp(a) the field includes pelacarsen at Novartis, olpasiran at Amgen, zerlasiran at Silence and lepodisiran at Lilly; in APOC3 it includes olezarsen at Ionis and plozasiran at Arrowhead.

The competitive framing that matters is not who lowers LDL-C the most. It is who can credibly claim years of effect from one administration with a safety profile acceptable in a preventive setting. Verve’s editing approach reaches durability by making a permanent change; the siRNAs reach safety and reversibility by dosing repeatedly. Scribe’s entire positioning is the claim that it can have both at once. Nobody has yet shown that in humans, including Scribe.

10Q2 2026 financial scorecard

The six-month picture is the more useful one. Collaboration revenue of $4.151 million against $22.023 million, research and development of $20.128 million against $29.876 million, general and administrative of $5.844 million against $9.332 million, a net loss of $23.824 million against $13.328 million, and stock-based compensation of $2.197 million against $1.347 million. The loss widened even though spending fell, because the revenue that offset spending in 2025 was largely recognition of performance obligations that have since been satisfied.

Balance sheet at June 30, 2026: total assets $57.927 million, total liabilities $102.417 million, total stockholders’ deficit $164.846 million and accumulated deficit $181.531 million. Those liability and deficit figures are pre-IPO artefacts and are the single most misleading line to quote without context. Of the $102.4 million of liabilities, $37.8 million was the fair value of the Lilly note that converted to equity in July and $39.2 million was deferred revenue, a non-cash obligation to perform research rather than to pay anyone. There was no term loan, no credit facility and no conventional bank debt.

11Cash, runway and a capital structure that was rebuilt in ten days

Three transactions in the second half of July changed the company’s financial identity. On July 17 the 5.9218-for-1 reverse split took effect and the Eli Lilly convertible note, $30.0 million of principal issued in May 2023 at 8% and matured on May 11, 2026, automatically converted into approximately 1.05 million shares together with roughly $7.8 million of accrued interest, at a conversion price of $35.83, with no cash payment required. The 10-Q does not publish the exact issued figure; the 1,054,828 shares that appear in the anti-dilutive table are the count computed at June 30, 2026. On July 23 the IPO priced at $15.00, the high end of the range and upsized. On July 27 the offering closed with the underwriters’ option fully exercised, at 9,867,000 shares in total, alongside 500,000 shares purchased for $7.5 million by Aventis and Genzyme, both Sanofi entities.

Aggregate gross proceeds were approximately $155.5 million and net proceeds approximately $140.7 million in the 10-Q and approximately $140.6 million in the second-quarter release, after roughly $14.8 million of underwriting discounts, commissions and expenses. Leerink Partners and Goldman Sachs acted as representatives of the underwriters, with Guggenheim Securities and Wells Fargo Securities also on the cover of the prospectus.

The prospectus allocation is specific: approximately $30 to $35 million to take STX-1150 through the ongoing Phase 1, approximately $15 to $20 million each for STX-1400 and STX-1200 through IND-enabling work, manufacturing, clinical initiation and initial readouts, approximately $20 to $25 million for the wider pipeline and platform, and the remainder for working capital. The stated runway on that plan reaches into the first half of 2029.

What remains on the dilution side is not debt but equity mechanics. At June 30 there were 1,068,963 options outstanding at a weighted average exercise price of $7.24, plus 74,324 warrants. On July 23 the company granted options over 2,066,997 shares at the $15.00 IPO price to directors, officers and employees, some with market-based vesting, and added 2,420,000 shares to the 2026 plan. Against 18.9 million shares outstanding, that reserve is material and should be carried in any per-share thinking. No at-the-market programme or shelf was disclosed in the 10-Q.

12Collaborations: what Sanofi and Lilly have actually paid

The two milestone figures for the Lilly agreement are not the same in the two documents, and neither filing reconciles them. The prospectus figures describe the agreement as signed; the 10-Q language describes what the company is currently eligible to receive. The safe reading is that some milestones have been earned or have lapsed since 2023, and that the remaining opportunity is the smaller number. Anyone modelling this line should use the 10-Q figure and treat the prospectus figure as the historical contract headline.

Cumulatively the company states it has received over $180 million in upfront, milestone and expense reimbursement payments from its collaborations. That is real validation and real cash, and it explains how a company with no product got to a first-in-human trial without an enormous venture burn. It is also, on the current run rate, a shrinking contribution: the revenue recognised in the first half of 2026 came entirely from the Prevail agreement, which also accounts for 96% of receivables, while the 2023 Sanofi agreement contributed no revenue at all in 2026, and total deferred revenue stood at $39.173 million at June 30 against $41.023 million at year-end.

Two other licences sit underneath the platform and are easy to overlook. The CasX and CasY patent rights are exclusively in-licensed from the Regents of the University of California under an agreement dated September 2018 and amended in 2020, carrying up to $27.7 million of development and regulatory milestones, up to $3.6 million of commercial milestones and low-to-mid single-digit royalties. The lipid nanoparticle used in STX-1150 is licensed from Acuitas Therapeutics, with up to $20.0 million of development and $40.0 million of commercial milestones per product; $2.0 million was recognised in research expense in the second quarter and $2.1 million of accrued licence fees sit on the balance sheet.

13Ownership, float and the January 2027 lock-up

The register is concentrated. On the prospectus table dated July 10, 2026, entities affiliated with Andreessen Horowitz held 32.24% before the offering and 15.61% after; Eli Lilly held 12.37% and 5.99%, a position created entirely by the note conversion; Avoro entities 8.18% and 3.96%; co-founder Jennifer Doudna 7.08% and 3.43%; co-founder Brett Staahl 6.34% and 3.10%; chief executive Benjamin Oakes 7.61% and 3.72%; OrbiMed 4.09% and 1.98%. All executive officers and directors as a group held 21.81% before and 10.78% after.

With 18.93 million shares outstanding and roughly 10.4 million of those sold in the offering and private placement, the tradeable float is small, and the trading data reflects it. Since listing, daily volume has run from roughly 50,000 to 990,000 shares, and the price has moved between a $18.06 close on July 28 and a $37.25 close on August 24 within five weeks. That is a stock where a modest change in demand produces a large price move in either direction.

14Management, board and the Doudna question

Benjamin Oakes, 37, co-founded the company in 2017 and serves as president, chief executive and chairman of the board. He holds a doctorate from UC Berkeley and was previously an entrepreneurial fellow at the Innovative Genomics Institute. Svetlana Lucas, chief business officer since 2019, became chief operating officer in July 2026; her background includes business development roles at Tizona, Amgen and Onyx, and she sits on the boards of Jasper Therapeutics and aTyr Pharma. David Parrot has been chief financial officer since December 2021, previously head of West Coast life science investment banking at Barclays. Those three are the only executive officers named in the prospectus; no chief scientific or chief medical officer is listed.

The board includes James Watson of Andreessen Horowitz, Behzad Aghazadeh of Avoro, Carl Gordon of OrbiMed and Joshua Bleharski as lead independent director. The combination of a founder chief executive who also chairs the board, with three of the remaining directors representing the largest shareholders, concentrates both economic and governance influence in the pre-IPO investor group. That is normal for a company at this stage; it is also the reason the lead independent director role and the eventual arrival of independent clinical expertise on the board are worth watching.

Jennifer Doudna’s role is frequently overstated in coverage of the stock, so it is worth stating precisely. She is a co-founder and the company’s most visible scientific name, and she held 7.08% before the offering. The prospectus is explicit that there are no plans for her to become an employee or a director following the offering, and that she is expected to continue under a consulting agreement. The Nobel association is real; day-to-day operational involvement is not what the filing describes. The third co-founder, David Savage, is at UC Berkeley, and Brett Staahl serves as vice president of external innovation and is principal investigator on the CIRM-funded programmes.

As of June 30, 2026 the company had 82 full-time employees, down from 89 at March 31 as reported in the prospectus, where 69 were in research and development and 44 held doctorates. The 10-Q refers to a prior-year reduction in force as a driver of lower salary costs without disclosing its date or size. The company reports as an emerging growth company and a smaller reporting company, which permits reduced disclosure and exempts it from the auditor attestation on internal control required by Section 404(b).

15Intellectual property and legal position

As of March 31, 2026 the estate comprised 12 issued US patents, 21 pending US non-provisional applications, 16 issued foreign patents, 194 pending foreign applications and two pending international applications under the Patent Cooperation Treaty, with expected expiry between 2037 and 2045 before any adjustment or extension. The foundational CasX and CasY rights are in-licensed exclusively from the University of California for all human diseases, excluding infectious viral diseases and diagnostics.

Two features distinguish this position from the older CRISPR companies. First, no Broad Institute licence dependency is stated, because the platform is built on CasX rather than on Cas9, which sidesteps the long-running Cas9 interference dispute. Second, both the prospectus and the 10-Q state that the company is not party to material legal proceedings. For a CRISPR company, an uncontested patent position is a genuine, if unglamorous, asset.

16Valuation, analyst coverage and the Merlintrader Health Score

Using the September 2, 2026 close of $30.90 and the 18,931,680 shares reported on the August 31, 2026 Form 10-Q cover gives a basic equity value of approximately $585 million. Subtracting roughly $183.6 million of pro forma cash implies the market is attributing something on the order of $400 million to a platform whose lead asset has no human efficacy data. This is a basic-share calculation, not enterprise value and not fully diluted; the option pool granted at listing and the 2026 plan reserve would both raise the denominator.

Conventional valuation methods do not apply here, and pretending otherwise would be false precision. There is no revenue to multiply beyond collaboration recognition that is running down, no approved product, and no risk-adjusted commercial model that would not be dominated by an assumption about human durability that nobody can yet make. What can be said is what the price implies: at this level the market is paying for the platform thesis rather than discounting it heavily, which is a different starting point from the average pre-data biotech.

On analyst coverage. Following the customary quiet period, several banks initiated in the second half of August 2026. Financial data providers report initiations from Leerink Partners at Outperform with a $37 target and Wells Fargo at Overweight with a $60 target on August 18, from Guggenheim at Buy with a $50 target, from H.C. Wainwright at Buy with a $60 target at the end of August, and coverage from Goldman Sachs. These figures are reported by data aggregators and have not been verified against the original research notes; three of the initiating firms were underwriters on the offering, which is normal practice and is also the reason such initiations are not independent evidence. Price targets are not used in this hub as a valuation input.

17Catalyst calendar and monitoring windows

Hard dates are separated from company targets throughout. Note what is absent: there is no PDUFA date, no advisory committee, no regulatory decision and no approval pathway event anywhere on this calendar, and there will not be for years. For a reader used to trading biotech around binary regulatory dates, this is a different kind of name.

18Scenario framework: no price targets, no false precision

No numerical price targets appear in this hub. Every input that would drive one, the depth of effect, its duration, the deliverable dose, the eventual label and the size of the preventive population, is unknown, and a model built on unknowns produces a number that looks like analysis and is not.

19Ranked risks and the falsifiable checklist

• 1. Durability may not translate. The company’s own risk factor states that epigenetic repression achieved in preclinical models may diminish over time in humans through chromatin remodelling or cellular turnover, potentially requiring re-dosing, and that durable silencing over multi-year horizons has not been conclusively demonstrated in human clinical studies. This is the thesis risk, not a footnote.
• 2. Single clinical asset. One molecule in humans, one target, one delivery system. A safety event in the Phase 1 would not be contained to STX-1150, because the LNP and the platform are shared with the programmes behind it.
• 3. Immunogenicity and re-dosing. The prospectus notes that immunogenicity may cause adverse effects and could prevent reapplication of the delivery method. For a therapy whose value depends on either lasting for years or being repeatable, that is a specific and material concern.
• 4. Timing risk on a guided window. First-half 2027 is company guidance, dependent on enrolment across two countries with one site currently active. Registry primary completion is December 2028.
• 5. Market structure. A small float, low message volume, wide daily ranges and a January 19, 2027 lock-up expiry sitting just before the guided readout.
• 6. Revenue decline. Collaboration revenue fell from $22.0M to $4.2M year over year in the first half, and every dollar recognised in 2026 came from a single agreement, the Prevail licence.
• 7. Competition and reimbursement. Approved chronic PCSK9 therapies, oral inhibitors in late development, and better-capitalised genetic-medicine rivals in all three of Scribe’s lipid targets.
• 8. Governance concentration. Founder chief executive who also chairs the board, three directors representing large pre-IPO holders, no chief medical officer named in the prospectus.
• 9. Dilution from equity plans. Options over 2,066,997 shares granted at listing plus a 2,420,000-share plan addition, against 18.93 million shares outstanding.

20Bottom line: a funded platform waiting for its first fact

Scribe Therapeutics is one of the cleaner examples of a company whose financial risk and scientific risk have been fully separated. The financial side is, for now, settled: roughly $183.6 million of pro forma cash, no debt after the Lilly note converted, non-dilutive CIRM money carrying two preclinical programmes, two large-pharma collaborations that have paid more than $180 million to date, and a stated runway into the first half of 2029 that reaches well past the first human readout. Very few companies at this stage of development have that.

The scientific side is entirely unsettled, and the company says so in its own filings. Epigenetic silencing has never been shown to last for years in a human being. The primate data are striking and the target is as validated as targets get, but the specific claim that makes this equity interesting, one dose and years of effect without touching the DNA, has exactly zero human observations behind it as of September 3, 2026.

That combination defines what kind of name this is. It is not a catalyst trade, because there is no dated regulatory event. It is not a valuation case, because there is nothing to value against. It is a platform hypothesis with a funded runway and a guided data window in the first half of 2027, preceded by a lock-up expiry on January 19, 2027 that falls at the start of that window, into a small float. Everything worth knowing about the science arrives on one date that has not been set yet.