Larger proteins may fold faster, organoids advance venom production research

A new rule of thumb is taking shape in protein folding

The surprising lesson from a recent protein-folding analysis is that bigger does not necessarily mean slower. In measurements led by researchers at the National Institute of Diabetes and Digestive and Kidney Diseases in Bethesda, Maryland, eight two-state single-domain proteins folded through their transition paths in about 0.7 to 4 microseconds, and the folding-barrier diffusion coefficient increased with protein length. That points to a more efficient crossing of the folding barrier in larger proteins, not because they are simpler, but because their native contacts may form more cooperatively.

That finding matters because it shifts the conversation from raw size to landscape design. If cooperative native-contact formation can reduce energy-landscape roughness, then the folding problem starts to look less like a race across a chaotic terrain and more like a route that biology has optimized for the most direct crossing point. The implication is not that all large proteins fold quickly, but that the old intuition about complexity slowing everything down is too blunt for what modern single-molecule methods can now see.

What the microsecond measurements actually change

The scale here is the story. Transition-path times in the microsecond range mean that the decisive part of folding happens extraordinarily fast, even for proteins that are not tiny. Hoi Sung Chung and William A. Eaton’s measurements, as discussed in the recent analysis, show that the barrier-crossing step is the key bottleneck, and that bottleneck appears to become more efficient as the proteins get longer within this set of eight examples.

That is a useful correction for anyone trying to model folding from first principles or build predictive tools for protein design. The data suggest that “larger” is not a proxy for “harder to fold” once native interactions are arranged to cooperate. Instead, the key variable may be how the chain organizes its contacts at the barrier, which is exactly the sort of detail that matters for computational design, protein engineering, and any pipeline that depends on getting a stable, functional structure on the first try.

The comparison with an engineered benchmark sharpens that point. The naturally occurring proteins in the study crossed the transition path faster than an engineered protein benchmark that took 13 microseconds, suggesting evolution may have tuned folding landscapes for rapid barrier crossing. That does not make engineered proteins inferior, but it does highlight how much room there still is to learn from biology’s own optimization strategies.

Why this matters for protein design and biotech pipelines

For protein engineers, the message is practical: sequence design is no longer just about stability at the endpoint. It is also about how a molecule navigates the route there. Cooperative native-contact formation, barrier roughness, and transition-path speed are becoming design variables, not just descriptive terms in a paper.

That has downstream consequences for industrial and biomedical workflows:

• Faster, more reliable folding can reduce aggregation problems in expression systems.
• Better barrier crossing can improve yield and reproducibility in therapeutic protein development.
• More predictive folding models can shorten the loop between computational design and wet-lab validation.

In other words, the field is moving from asking whether a protein can fold to asking how the folding pathway itself can be engineered. That is a bigger shift than a single microsecond number might suggest.

Venom organoids turn a hard-to-handle biological supply into a lab system

The same broad trend toward controllable biology shows up in snake venom gland organoids. First reported in Cell in 2019, and later described in a 2020 Nature Protocols paper, these organoids preserve the cellular heterogeneity of the venom gland while letting researchers influence cell type composition through changes in media composition. That combination is unusually powerful: it keeps the biology rich while giving the lab a lever to shape it.

The practical appeal is obvious. The Leiden protocol notes that manually milking snakes remains the most common way to obtain venom, but it is labor-intensive and dangerous. Venom is also notoriously variable, which complicates both antivenom production and drug discovery. A lab-grown system that can be expanded indefinitely, harvested repeatedly, and still produce active toxins from multiple snake species changes the economics and the safety profile of venom research.

At the Hubrecht Institute, with collaborators linked to Oncode Institute and Princess Máxima Center in The Netherlands, the work has been framed as more than a technical curiosity. It is a platform for generating biologically active venom components in a controlled setting, which opens the door to more consistent toxin libraries, cleaner comparisons across species, and faster iteration on antivenom targets. Reviews and institutional writeups have also emphasized the therapeutic angle, where venom-derived molecules can become starting points for drug discovery rather than just hazards to contain.

A broader shift: biology is becoming more predictive and more engineerable

Put the folding work and the organoid work side by side, and the pattern becomes clear. In both cases, researchers are moving away from static descriptions of biology and toward systems that can be measured, modeled, and shaped with much finer control. The folding study suggests that even very small dynamical events can reveal design principles hidden inside proteins. The organoid system shows that a difficult tissue can be turned into a repeatable production platform without losing its essential complexity.

That is why these updates matter beyond their immediate niches. A 2026 Nature review describes organoids as important tools for studying human development and disease, for regenerative medicine, and for drug screening. Add that to the new folding results, and the direction of travel is hard to miss: biology is becoming a field where speed, scale, and architecture can all be treated as engineering variables.

For biotech pipelines, that is the real takeaway. Protein science is getting more predictive at the molecular level, and organoid systems are making previously messy biological processes more manufacturable. The result is a research landscape where fundamental insight and practical application are increasingly built from the same materials.