👋 Hey, let’s get into it. Anthropic has officially entered its wet-lab era.

The AI company confirmed it has established a biology lab in the Bay Area, moving some of its life-sciences work beyond models and predictions. Anthropic later clarified that the facility is not specifically for drug discovery, so this is not Claude Pharmaceuticals. At least not yet.

Still, there is only so far you can get in biology with a computer. Eventually, somebody has to run the experiment. Anthropic now appears to want a seat at that bench too.

📉 Novo’s Growth Plan Sends Shares Down as Much as 9%
Novo laid out plans to launch more than five multi-blockbusters by 2030, reach 60 million patients, and generate more than DKK150 billion ($23B) in risk-adjusted annual pipeline sales in 2035. Investors were unmoved. Shares fell as much as 9% intraday Monday as management faced questions about pricing, competition, and replacing semaglutide revenue before its early-2030s patent expirations. A week after shortening its everyday brand to Novo, the company supplied plenty of long-term ambition. But a shorter name and a longer slide deck cannot close the execution gap by themselves.

☢️ Telix Agrees to Buy ITM in Deal Worth Up to $2.35B
Telix agreed to buy German radiopharma company ITM in a deal valued at $1.65 billion on a cash-free, debt-free basis, plus up to $700 million in conditional payments. Importantly, this is not a $1.65 billion cash check. About $1.25 billion of seller consideration is expected in Telix shares. The acquisition adds investigational therapy ITM-11 and an established isotope business. ITM supplies lutetium-177 and reported $273 million in 2025 revenue. In radiopharma, finding a promising drug is only half the battle. You also need a dependable isotope supply. Telix is buying both.

🎗️ FDA Approves Lilly’s Inluriyo-Verzenio Breast Cancer Combo
The FDA approved Lilly’s Inluriyo plus Verzenio for adults with ER-positive, HER2-negative, ESR1-mutated advanced or metastatic breast cancer whose disease progressed after endocrine therapy. In EMBER-3, the combination extended median progression-free survival to 11.1 months versus 5.5 with Inluriyo alone, reducing the risk of progression or death by 47%. The decision gives Lilly a two-drug option combining estrogen-receptor degradation with CDK4/6 inhibition. That’s a meaningful expansion in the competitive oral estrogen-receptor-degrader market.

💊 Novo Signs Oral-Drug Discovery Deal Worth Up to $1.4B
In other Novo news, the company signed a multi-target partnership with Orbis Medicines to discover oral macrocyclic drugs against cardiometabolic targets. Macrocycles are ring-shaped molecules that may reach targets conventional small molecules struggle to engage. The deal could be worth up to $1.4 billion in upfront and milestone payments, plus royalties and an undisclosed strategic investment. The upfront amount and targets remain secret. The appeal is easy to understand: injectable medicines work, but many patients would rather take a pill. The caveat is even simpler. This is early discovery, not an oral Wegovy waiting in the wings.

Stop hoping the algorithm finds your audience. They’re already here.

BioNucleus Briefing has 1,600+ subscriptions across its email and LinkedIn newsletters, including scientists, founders, investors, operators, and other biotech and pharma professionals.

We feature one exclusive sponsor per issue, giving relevant tools, software, services, events, and other life-science brands a direct way to reach our readers.

Think your company could be a fit? Reply to this email and let’s talk.

Alright, back to the newsletter.

🤝 Roche Signs Dualitas Antibody Deal Worth Up to $1B
Roche will pay Dualitas $36.5 million upfront to discover bispecific antibodies for immune diseases, with milestones potentially pushing the deal to $1 billion.

🧬 Ultragenyx Wins FDA Approval for First Sanfilippo Type A Treatment
The FDA approved Fayuvi, a one-time intravenous gene therapy and the first treatment for children with Sanfilippo syndrome type A, a progressive neurodegenerative disorder.

🌏 U.S. Reportedly Weighs Allowing Most China Drug-Licensing Deals
U.S. officials are reportedly considering rules allowing most China drug-licensing deals, with exceptions for biotechnology that could potentially be weaponized.

⚠️ Xenon Pauses Psychiatric Trial Enrollment After Safety Events
Xenon voluntarily paused new enrollment in depression and bipolar studies after neuropsychiatric events, sending shares down about 26% premarket while epilepsy development continued.

💰 Lisata Acquires Marea Alongside $225M Financing Agreement
Lisata acquired Marea and signed a $225 million private placement, funding a cardioendocrine pivot and projected runway into 2028 for the combined company.

👁️ Beacon’s Gene Therapy Wins Pivotal Inherited-Blindness Trial
Beacon’s laru-zova improved low-light vision in a Phase 2/3 X-linked retinitis pigmentosa trial, supporting a planned rolling FDA filing later this year.

🧩 Novo Buys Three Non-Incretin Obesity Programs From Kallyope
Novo agreed to buy three early-stage obesity programs from Kallyope, adding non-incretin mechanisms without disclosing detailed targets or deal terms.

🇨🇳 China Sets New Drug-Innovation Targets for 2030
China’s new pharmaceutical plan aims for the country to originate at least 25% of global first-in-class drugs by 2030, among broader industry targets.

Designing a promising clinical trial is hard. Making it practical for patients to join can be harder. At DPHARM, executives from Merck, Gilead, Lilly, and BeOne described a recruitment system strained by saturated sites, complicated workflows, and difficulty reaching patients who reflect the people a drug may ultimately serve.

A small May survey, sponsored by trial-access company myTomorrows and reported by Fierce, shows what that strain can look like. Among 100 U.S. site employees, 59% said process inefficiencies reduced the number of eligible patients recruited. More than half used at least four prescreening tools, while 27% spent over 20 hours a week reviewing referrals. Nothing says efficiency like opening tab number 37 to find out whether a patient belongs in tab number 12.

Panelists pointed to community sites, patient input during trial design, and electronic-health-record matching as ways to widen access and identify potential participants earlier. FDA guidance also allows decentralized elements such as telehealth, home visits, and local providers. These approaches can move parts of a trial closer to patients without turning every study into a fully remote one.

But identifying a possible match is only the beginning. Site staff still must confirm eligibility, support informed consent, assess safety, and determine whether appointments, travel, childcare, and time away from work make participation realistic. A better dashboard may generate more names, but it does not create coordinator hours or make a burdensome protocol easier to live with.

For sponsors, slow recruitment means delayed timelines and higher costs. For sites, it means more administrative work. For patients, a trial can remain theoretically available but practically unreachable. Recruitment improves when technology removes work and participation barriers, not when it adds another login.

This is the number of times higher lung-cancer risk was among never-smokers carrying the rare inherited EGFR T790M mutation compared with never-smokers without it. The variant appears in only about one in 15,000 people, and absolute lifetime risk remains unclear.

Here’s what biotech Redditors are talking about:

😬 Biotech Interview Questions Get Weird
Biotech Redditors shared questions that blindsided them, from choosing an element to naming the problem they’d solve with unlimited resources. Moral of the story: you can prepare to discuss assay troubleshooting, only to be asked for your elemental alter ego. Emergency answer: “I’m like sodium (Na). Highly sensitive, extremely reactive, and if dropped into water, I will literally explode.” I can’t guarantee it’ll land you the job, but the interviewer will remember you. Reply with the weirdest question you’ve received.

🤖 Your AI Assistant Isn’t the Interview Candidate
A hiring manager described a technical interview that turned weird when a candidate paused for an awkwardly long time and then delivered rambling answers to questions about HPLC troubleshooting. The interviewer suspected AI was supplying answers offscreen. The consensus on Reddit is that using AI beforehand to practice is perfectly okay and, in fact, recommended. Letting it become your invisible co-applicant is another matter. The interview is supposed to show what you know, not what your chatbot knows.

🤖 FDA Clears Neptune’s Robotic Colonoscopy System
The FDA cleared Neptune Medical’s physician-controlled Triton 1 robotic endoscopy system for diagnostic and therapeutic procedures throughout the colon, bringing robotic control into gastrointestinal care.

📄 Paper2Agent Turns Research Papers Into AI Research Tools
Paper2Agent converted 74 of 100 computational-biology papers into reusable AI-accessible workflows, helping researchers run published methods through natural-language instructions and inspect the results.

🐭 Artelo’s Cannabinoid Drug Adds to GLP-1 Weight Loss in Mice
Artelo’s ART27.13 roughly doubled weight loss alongside semaglutide in obese mice, but no human obesity benefit has been shown yet…mice win again.

🧩 ARPA-H Launches SPECTRA for Autism Research
ARPA-H launched SPECTRA, a program seeking to integrate biological, clinical, behavioral, and real-world data for more individualized autism assessment and support.

🍷 Goa Launches With $15.5M to Build the “Narcan for Alcohol”
Goa Therapeutics emerged from stealth with $15.5 million and GOA26, an experimental drug designed to accelerate alcohol clearance. The company’s pitch is “Narcan for alcohol,” initially targeting emergency care for intoxicated trauma patients who may need urgent treatment. Goa reported that treated pigs had 61.3% lower blood-alcohol concentrations than controls 20 minutes after dosing. An IND filing is planned by year-end, followed by Phase 1 in 2027. For now, put away the brunch plans. This is not a hangover cure just yet.

🧪 September 22, 1901: The Man Who Put Cancer on a Hormonal Diet Is Born
On this day 125 years ago, Charles Brenton Huggins was born in Halifax, Nova Scotia. Four decades later, he helped show that prostate tumors were not growing entirely on autopilot. They depended on androgen signals, and cutting off that supply could cause advanced tumors to regress. The discovery helped establish the idea that cancer’s biological dependencies could become therapeutic targets. Huggins shared the 1966 Nobel Prize in Physiology or Medicine, and endocrine therapy remains a cornerstone of cancer care today.

Thinking back to Goa’s plan to build the “Narcan for alcohol,” I have one important question if GOA26 reaches human clinical trials: Does enrolling entitle one to free beer? Or is it more of a BYOB situation? Asking for a friend.

We only ask the important questions here at BioNucleus.

That’s all for today. See you Thursday for the next issue. 👋

✍️ Today’s email was brought to you by Josh Martin. Connect with me on LinkedIn.

If you’re enjoying BioNucleus, share it with a friend or coworker who’d get value from it. We’re a small operation, and every share makes a bigger difference than you think.

Was this email forwarded to you? Sign up here 👇

Subscribe for free

Follow us on social and stay one step ahead