One Target, Three Development Strategies

Choosing a target does not yet choose a product. A team pursuing PCSK9 can intercept the circulating protein or reduce its production in the liver. That choice changes what must be designed, where it must go, and which experiment will show that the route is working.

We use three historical published examples to examine that choice. The comparison concerns biological action, delivery, and the next development decision. It does not imply that Crane carried out these studies or offers an end-to-end service for every modality.

The PCSK9-antibody structure in PDB 3H42 shows one extracellular blocking mechanism. The ORION-10 and ORION-11 trials measured a different route: GalNAc-conjugated siRNA delivered to the liver to reduce PCSK9 synthesis. The phase 2b MK-0616 study measured yet another: an oral macrocyclic peptide that inhibits PCSK9.

These sources illustrate distinct routes. They are not a head-to-head efficacy comparison or a review of the current treatment landscape. They are also different design tasks. An antibody campaign must resolve epitope, paratope, specificity, expression, and developability. An siRNA campaign depends on sequence, off-target risk, intracellular persistence, and tissue delivery. A macrocycle must solve binding geometry together with chemical stability, permeability, and oral exposure.

A broad model catalogue cannot resolve the route decision. A team first needs a credible account of how the molecule will reach its site of action and produce the intended effect. The route comparison above makes those assumptions explicit before a design campaign scales.

Crane's contribution here is protein and peptide design: assessing interfaces and formats, comparing structural hypotheses, and selecting candidates for experimental evaluation. A route that depends on RNA delivery or another chemistry needs the corresponding specialist expertise and evidence. Representing that chemistry in a model does not establish a validated development workflow.

For a binding program, the next decisions are the target's relevant state and the surface that supports the mechanism. For an enzyme program, the starting point is the chemistry the protein must support. These are different briefs, even when they draw on some of the same models.