Biopharmaceuticals

This page covers what all protein therapeutics have in common because a cell, not a chemist, assembles them. It deliberately does not cover the individual platforms: conjugation chemistry, bispecific engineering, oligonucleotide pharmacology and cell products each have their own page, and the mechanism of a particular drug class belongs there.

The process is the product

A small molecule is fully specified by a structural formula; two batches made by different routes are the same substance. A therapeutic protein is not. Its amino-acid sequence is only the primary structure; activity requires correct folding, correct disulfide pairing, and a set of post-translational modifications the host cell decides on. Glycosylation is the clearest case: the glycan attached at the conserved asparagine in an IgG Fc is not a decoration. Afucosylation raises FcγRIIIa affinity and antibody-dependent cytotoxicity by more than an order of magnitude; terminal sialic acid extends serum half-life; exposed mannose residues shorten it through mannose-receptor clearance. Cell line, feed, dissolved oxygen and pH all shift the glycan profile, so a change in the bioreactor changes the drug’s pharmacology.

This is why regulators treat manufacturing changes as regulated events, and why a biosimilar is legally a different category from a generic. It is demonstrated to be highly similar through comparative analytics, functional assays and clinical pharmacology rather than by proving chemical identity, because identity is not achievable.

Why these drugs are injected

Two properties rule out the oral route for almost all of them. Proteins are substrates for gastric and pancreatic proteases, and even intact they are far too large and too polar to cross the intestinal epithelium in useful quantity — oral bioavailability without an absorption enhancer sits at a fraction of a percent. The exceptions prove the constraint: oral semaglutide is formulated with salcaprozate sodium to buy local absorption in the stomach, and still requires a much larger dose than the injected form.

Circulating half-life follows from a second mechanism. Antibodies and albumin persist for weeks because FcRn in endothelial and myeloid cells binds them in the acidified endosome and recycles them to the surface instead of routing them to the lysosome. Almost every half-life-extension strategy in the field — Fc fusion, albumin binding, engineered pH-dependent Fc variants — is an attempt to enter that recycling loop. Proteins outside it, such as most cytokines and enzymes, clear in hours.

The ceilings

Immunogenicity is the one that never fully goes away. Anti-drug antibodies can neutralise activity or accelerate clearance, and their risk is driven less by sequence humanisation than by aggregates, subvisible particles and product-related impurities, which is why aggregation control dominates formulation work. Stability sets the second ceiling: proteins denature, deamidate, oxidise and aggregate, so most products need cold chain and none tolerates the thermal abuse a tablet does. And dose-response is not the small-molecule kind — target-mediated drug disposition means the target itself is a clearance route, so exposure can rise non-linearly once receptor binding saturates.

See also the technology article at /technology/biopharmaceuticals-biomedicines/.