On August 24, 2026, Hanmi Pharm announced an exclusive licensing agreement with Genentech (a Roche Group member) for HM17321, an investigational urocortin-2 (UCN2) analog peptide for obesity and related conditions. Genentech paid $190 million upfront against a total deal value of up to roughly $2.3 billion, including milestones and tiered royalties, for worldwide rights excluding South Korea. The striking part is the timing: Genentech is taking on development starting at Phase 2, meaning the entire deal was struck with only preclinical efficacy data and an ongoing Phase 1 safety study — no human efficacy results — behind it.
This article explains what UCN2 signaling is, why a major pharma company paid a pre-Phase-2 price for a muscle-preservation mechanism, how HM17321 could theoretically pair with GLP-1 drugs in fixed-dose combinations, and — critically — where the evidence actually stands today. Every efficacy claim about HM17321 to date is preclinical.
The deal, at a glance
The core terms come from Hanmi's own announcement, distributed via PR Newswire and independently corroborated by HealthCare Middle East & Africa:
| Deal element | Detail |
|---|---|
| Asset | HM17321, a proprietary UCN2 (urocortin-2) analog peptide |
| Mechanism | Non-incretin; described as potential first-in-class |
| Scope | Exclusive worldwide license, excluding South Korea |
| Upfront | $190 million |
| Total potential value | ~$2.3 billion (development, regulatory, commercial milestones) |
| Royalties | Tiered, on future net sales |
| Current stage | Phase 1 (NCT07219589), safety/tolerability/PK/PD in healthy volunteers and people with obesity |
| Division of work | Hanmi completes Phase 1; Genentech takes over from Phase 2 |
Pharmaceutical Technology characterized it as Roche's joint-largest deal of 2026 so far — a notable distinction for an asset with no Phase 2 data. The licensing covers development, manufacturing, and commercialization for obesity and associated conditions including type 2 diabetes and cardiovascular diseases.
What UCN2 signaling actually is
Urocortin-2 is a naturally occurring signaling peptide in the corticotropin-releasing factor (CRF) family. It binds to a receptor called CRHR2 (corticotropin-releasing hormone receptor 2), which, unlike the gut- and brain-centric receptors targeted by GLP-1 drugs, is highly expressed in skeletal muscle. According to TechTimes' coverage, activating CRHR2 in skeletal muscle triggers a cascade that includes the mTOR protein-synthesis pathway — one of the master intracellular switches controlling muscle protein synthesis — while UCN2 signaling in adipose tissue promotes lipolysis (fat breakdown) and suppresses lipogenesis (fat storage).
That dual profile — an anabolic signal to muscle plus a lipolytic signal to fat — is the entire thesis. Where GLP-1 receptor agonists like semaglutide and tirzepatide work primarily by suppressing appetite and slowing gastric emptying, a UCN2 analog would, in theory, act directly on body composition: flip muscle's own growth machinery on while encouraging fat catabolism, without manipulating hunger at all. Pharmaceutical Technology adds that UCN2 has long been known to reduce rates of muscle atrophy, which is why the pathway has drawn attention for weight-loss applications.
Two practical notes matter here. First, HM17321 is a peptide delivered by subcutaneous injection — the same route as GLP-1 drugs — which makes chronic use and fixed-dose combinations more feasible than infusion-based muscle-preserving approaches such as antibody myostatin inhibitors. Second, per the company press release, the molecule came out of Hanmi's proprietary HARP (AI-driven) discovery platform, illustrating how computational peptide discovery is increasingly feeding big pharma's metabolic pipelines — a pattern we've also followed in coverage of AI-adjacent pipeline bets like Fractyl's Rejuva gene therapy and peptide-based cholesterol drugs like enlicitide.
Evidence boundary: the CRHR2-in-muscle mechanism and mTOR activation are established biology; that HM17321 produces meaningful fat-selective weight loss with preserved lean mass in humans is entirely unproven. The strongest data so far are preclinical — including, per TechTimes, fat-selective reduction with preserved lean mass in Rhesus monkeys presented at EASD 2025, and body-composition improvements in animal models that improved further when combined with semaglutide, presented at ADA 2025.
Why big pharma is paying pre-Phase-2 prices for muscle preservation
The strategic logic rests on the single biggest clinical complaint about GLP-1 therapy: the weight lost is not all fat. Per TechTimes, published data indicate that lean body mass accounts for roughly 25–39% of total weight lost on GLP-1 therapies — approximately 25% with tirzepatide and up to about 39% with semaglutide among patients losing more than 15% of body weight. That range is actively debated in the literature.
The debate spilled into the open at the ADA 2026 Scientific Sessions in New Orleans. Dr. Samuel Klein of Washington University argued that lean mass loss on GLP-1 drugs is proportional to total weight loss — the same pattern seen with any major caloric restriction — and that no current evidence links it to frailty or clinically defined sarcopenia. Dr. Eric Ravussin of LSU's Pennington center countered that older patients with low muscle reserves face real clinical exposure. Both positions can be true: the mechanism is not GLP-1-specific catabolism, but a large energy deficit without simultaneous muscle anabolism will naturally draw on lean tissue — and for the substantial share of GLP-1 users over 60, or patients who cycle off therapy and regain weight (TechTimes cites data suggesting up to two-thirds discontinue within a year), that proportion can cross into clinically meaningful territory.
Against that backdrop, the deal structure tells you what Roche thinks it is buying. A $190 million upfront payment for a Phase 1 asset is a bet that body composition — not just scale weight — becomes the next axis of competition in obesity medicine, and that securing a first-mover position on the UCN2/CRHR2 pathway is worth buying before the Phase 2 readout reprices it. TechTimes noted that Denmark's Gubra started its own Phase 1/2 trial of a competing UCN2 analog, GUB-UCN2 (~188 participants, with muscle volume and function endpoints), in July 2026 — giving the Hanmi/Genentech program roughly an eight-month head start. Roche, in other words, is paying to own a scarce mechanism before the field crowds.
How HM17321 could pair with GLP-1s — and why it fits Roche's portfolio
The mechanistic appeal of a UCN2 analog in combination is orthogonality. GLP-1 drugs drive weight loss through reduced caloric intake; UCN2 analogs would, in theory, act on the composition of what is lost by engaging muscle anabolic signaling directly. Hanmi's press release states that as a peptide-based therapeutic, HM17321 "may have future potential for use in fixed-dose combination (FDC) products or combination regimens with incretin-based treatments" — and that preclinical studies showed weight-reduction improvements both as monotherapy and combined with GLP-1-based therapies.
This is portfolio construction, not replacement. Genentech's obesity pipeline already includes enicepatide, a cAMP-biased dual GLP-1/GIP agonist in Phase 3, and petrelintide, an amylin analog also in Phase 3, with a Phase 2 multi-arm trial of enicepatide/petrelintide fixed-dose combinations announced in mid-2026 (TechTimes). Roche's pipeline also includes emugrobart, an anti-myostatin antibody with ongoing obesity studies (Pharmaceutical Technology). HM17321 adds the one thing none of those assets provides: a non-incretin mechanism explicitly aimed at body composition, stackable with the incretin franchises Roche is already building. Roche business-development head Boris L. Zaïtra framed it exactly that way, describing "a differentiated approach to selectively reduce fat mass while improving both muscle mass and muscle function."
The fixed-dose-combination angle is where the peptide nature of HM17321 does real work. Two subcutaneously delivered peptides with complementary mechanisms can plausibly be co-formulated — the same logic behind many incretin combination programs — whereas an infused biologic cannot. Whether that combination actually improves outcomes will require dedicated clinical trials that have not yet been designed, let alone run.
What this does not mean
This is the section that matters most, because the headline numbers invite over-reading.
- It does not mean HM17321 works in humans. The Phase 1 trial (NCT07219589) measures safety, tolerability, pharmacokinetics, and pharmacodynamics — it is not designed to demonstrate whether HM17321 causes meaningful fat loss or muscle preservation at therapeutic doses in people.
- The $2.3 billion is not a valuation of proven efficacy. Roughly $2.11 billion of it is milestone payments contingent on clinical, regulatory, and commercial success. The committed, non-refundable amount is $190 million. If Phase 1 shows safety problems, the milestones never trigger.
- A press release is not peer review. All efficacy characterizations — "simultaneously promote weight loss and preserve lean body mass," improvements with GLP-1 combinations — are company claims about preclinical studies, not published human trial results.
- Animal and primate results have a mixed translation record. Primate physiology is a meaningful translational step, but metabolic medicine in particular has repeatedly seen compelling animal body-composition data fail at the human clinical stage. The fact that Eli Lilly previously culled a Phase 1 UCN2 heart-failure program in 2019 without explanation (Pharmaceutical Technology) is a useful corrective to the assumption that UCN2 biology converts smoothly into drugs.
- Nothing is approved, and nothing is available. HM17321 is not FDA-approved, not prescribed, not offered through telehealth, and should not be confused with research-grade "UCN2" peptides sold for laboratory use, which are not drug products and are not for human consumption.
- The deal is not a signal about any specific product's price, availability, or timing. It is an option on a hypothesis. Phase 2 results — Genentech's responsibility — are, per TechTimes' assessment, at minimum several years away.
One useful comparison for calibrating expectations: we recently covered BGE-102, an oral NLRP3 inhibitor that cut hsCRP roughly 85% in a Phase 1 trial — and even that striking Phase 1 biomarker result still left the hard work of Phase 2 efficacy entirely ahead. HM17321 has less human data than BGE-102 does. The upfront price reflects the scarcity of the mechanism and the buyer, not the maturity of the evidence.
The race and the record
Two pieces of context sharpen the picture. First, this is Hanmi's second major metabolic licensing deal in under three months: on June 1, 2026, Hanmi licensed sonefpeglutide, a long-acting GLP-2 analog, to Eli Lilly for up to $1.26 billion including a $75 million upfront (TechTimes). The Genentech deal more than doubles both figures and sets a new single-asset record for the company.
Second, the market backdrop explains the urgency. GlobalData's forecast, cited by Pharmaceutical Technology, projects the seven-major-market obesity drug market growing at a 32.3% CAGR to $173.5 billion by 2031. Approved weight-loss drugs overwhelmingly target GLP-1 and GIP receptors; a validated UCN2/CRHR2 mechanism would be the first dedicated pharmacological approach to body composition. That combination of market scale and mechanism scarcity is precisely why an eight-month clinical head start was worth nine figures.
What to watch next
For readers tracking the non-incretin frontier, the meaningful upcoming evidence is: (1) Phase 1 safety, tolerability, PK/PD data from NCT07219589, which Hanmi will complete before handing off; (2) any Phase 2 design Genentech discloses, particularly whether it includes body-composition endpoints like lean-mass preservation and muscle function rather than weight alone; and (3) Gubra's competing GUB-UCN2 Phase 1/2, which per TechTimes includes muscle-specific endpoints and could produce the field's first comparative signal. Until then, every claim about HM17321's promise should carry the same qualifier Roche's own deal terms imply: this is a bet on biology, placed before the biology has been tested in humans.
This article is for general informational purposes only and is not medical advice. Talk to a licensed clinician about personal medical decisions.
Frequently Asked Questions
What is HM17321? An investigational analog of urocortin-2 (UCN2), a peptide in the corticotropin-releasing factor family, discovered by Hanmi Pharm and described in its press release as a potential first-in-class, non-incretin obesity treatment designed to promote weight loss while preserving lean body mass.
Is HM17321 approved or available? No. It is in an ongoing Phase 1 trial evaluating safety, tolerability, pharmacokinetics, and pharmacodynamics in healthy volunteers and people with obesity. It is not approved, not prescribed, and not legally available for human use in any form.
Why did Genentech pay $190 million upfront before human efficacy data? The price reflects conviction in the UCN2/CRHR2 muscle-preservation hypothesis and a strategic move to secure the pathway ahead of competitors like Gubra before a Phase 2 readout reprices it. The bulk of the $2.3 billion headline value is contingent on clinical, regulatory, and commercial milestones.
Do GLP-1 drugs destroy muscle? Not selectively. Roughly 25–39% of weight lost on GLP-1 therapies is lean mass — largely proportional to the caloric deficit, not a unique destructive effect. Whether that loss is clinically harmful, particularly for older adults with low muscle reserves, is an active scientific debate.
Has any UCN2-targeting drug ever been approved? No. Eli Lilly previously discontinued a Phase 1 UCN2 heart-failure program in 2019 without disclosing its reasons, underscoring that early UCN2 programs have failed before.
How could HM17321 pair with GLP-1 drugs? Because the mechanisms are orthogonal, Hanmi has said the peptide may have future potential in fixed-dose combination products or combination regimens with incretin therapies. This is preclinical rationale only — no combination has been tested in humans.
Continue your research
Sources
- Hanmi Pharm, "Hanmi Pharm Signs Exclusive Licensing Deal with Genentech for Novel Obesity Therapy," PR Newswire via Morningstar, August 24, 2026 — https://www.morningstar.com/news/pr-newswire/20260824cn31969/hanmi-pharm-signs-exclusive-licensing-deal-with-genentech-for-novel-obesity-therapy
- Benjamin Eddie, "Genentech Pays $190M for Obesity Drug Built to Preserve Muscle as GLP-1 Drugs Erode It," TechTimes, August 24, 2026 — https://www.techtimes.com/articles/325371/20260824/genentech-pays-190m-obesity-drug-built-preserve-muscle-glp-1-drugs-erode.htm
- "Hanmi Pharm licenses obesity drug candidate HM17321 to Roche's Genentech in US$2.3B deal," HealthCare Middle East & Africa, August 25, 2026 — https://www.healthcaremea.com/2026/08/25/hanmi-pharm-licenses-obesity-drug-candidate-hm17321-to-roches-genentech-in-us2-3b-deal/
- Robert Barrie, "Roche outlays $2.3bn for Hanmi's non-GLP-1 obesity candidate," Pharmaceutical Technology, August 24, 2026 — https://www.pharmaceutical-technology.com/news/roche-outlays-2-3bn-for-hanmis-non-glp-1-obesity-candidate/