Semaglutide P29: a new production method

A 2026 study published in PLoS ONE reports a new method for making the semaglutide main-chain intermediate at yields that previous lab-scale strategies have not matched: 5.85 grams of purified P29 per litre of bacterial broth, at 98% purity^[1]. The work matters because semaglutide, sold as Ozempic, Rybelsus and Wegovy, is one of the highest-demand prescription medicines in the world, and the conventional manufacturing process has struggled to keep pace with that demand.

What P29 is and why it is the hard part

Semaglutide is a 31-amino-acid GLP-1 receptor agonist. Before the finished drug can be made, manufacturers need a clean supply of the 29-residue main chain, formally called Arg34GLP-1 (9-37) and known in the literature as P29^[1]. Getting P29 at scale is the rate-limiting step in the semi-recombinant process Novo Nordisk uses for the licensed medicine. The dominant industrial approach starts from a recombinant precursor expressed in bacteria, then uses chemical and enzymatic steps to trim it down to the right sequence. The problem is that bacterial expression tends to produce the precursor tangled in inclusion bodies, which need extensive renaturation before any downstream processing can work.

Chemical solid-phase peptide synthesis is an alternative, but P29 is long enough that synthetic yield drops and cost rises sharply as the chain length grows. The result is that both major routes face meaningful cost and scale constraints at the volumes the GLP-1 market now demands.

What the new strategy does differently

Qi and colleagues at the Institute of Biochemical Engineering, Beijing University of Chemical Technology, designed a fusion construct they call a SNAC-tagged enterokinase-cleavable fusion peptide^[1]. The construct contains the sequence GSHHWHHHSSGDDDDK, in which the GSHHW section is the SNAC motif and DDDDK is the enterokinase recognition sequence. Multiple P29 units are linked in tandem separated by this motif, so each bacterial cell carries several copies of the target sequence per expression cassette.

After expression in Escherichia coli BL21(DE3) with IPTG induction, the fusion protein is processed in two sequential cleavage steps. First, nickel-ion-assisted chemical cleavage runs at pH 8.6 for 40 hours to cut the SNAC tag. Second, enterokinase digestion at 37 degrees Celsius for 5 hours releases the P29 monomer from the residual fusion sequence^[1]. The authors report that the two-step design avoids the extensive renaturation that single-step inclusion-body approaches require and reduces the reliance on resin-based chromatographic purification.

The numbers

The headline productivity figure is 5.85 g of P29 per litre of fermentation broth^[1]. Purity after the cleavage and purification steps reached 98%. Secondary structure analysis of the purified product showed 59.8% alpha-helices, 4.5% beta-sheets, 19.4% beta-turns and 22.0% random coils, which the authors compare with the expected folding profile for a biologically active GLP-1 analogue.

The 5.85 g/L figure is the key comparison point. The authors state the result represents improved productivity relative to previously reported recombinant strategies for the same intermediate, though direct head-to-head data from those earlier processes are not reproduced in the PLoS ONE paper.

What this is not

This is a proof-of-concept study at lab scale, not a validated industrial process. The E. coli system the authors used is common in academic fermentation research; whether it transfers to the fermentor volumes and regulatory requirements of a commercial pharmaceutical facility is a separate question the paper does not address. P29 is also only the main-chain intermediate: the finished semaglutide molecule requires additional chemical steps to attach the fatty-acid side chain that gives the drug its long half-life^[2]. Those downstream steps are not covered in this study.

The work does not have any direct bearing on existing approved semaglutide products. Ozempic and Wegovy are manufactured by Novo Nordisk under a process reviewed by regulators, and neither the EMA nor the FDA has changed their assessment of those medicines as a result of this publication.

Why manufacturing research matters for patients

The GLP-1 drug shortage of 2023 and 2024 showed that even a medicine with strong clinical evidence can be unavailable at scale when manufacturing capacity lags behind prescribing volumes. Better recombinant production routes for semaglutide intermediates could, in principle, support biosimilar development or provide manufacturing redundancy as more countries move semaglutide into national formularies^[3]. Biosimilar regulatory pathways vary by jurisdiction; the EMA and FDA each have their own requirements for demonstrating comparability with the originator product.

For patients and prescribers, the immediate picture is unchanged: semaglutide remains a prescription medicine regulated under the same rules as before this paper. Country-by-country classification and access criteria are on the semaglutide regulation pages. If you are considering any semaglutide product, that conversation belongs with a clinician who knows your history.