ENDE
AEVORIANA

Molecules in ice: what cryo-electron microscopy really shows

Science · Feature

Cryo-electron microscopy makes the three-dimensional architecture of biological molecules accessible without requiring a crystal for every sample. Extremely rapid cooling traps them in thin, glass-like ice. A three-dimensional reconstruction emerges from many weak individual images. Between the raw image and the structural model, however, lie numerous selection decisions, checks and methodological limitations.

Why glass-like ice is crucial

Biological molecules usually function in an aqueous environment. This environment is difficult to preserve in the vacuum of an electron microscope, and the electron beam can damage the sample. In cryo-electron microscopy, a very thin layer of liquid is cooled so rapidly that, under suitable conditions, the water solidifies into a glass-like state rather than forming large ice crystals. This holds the molecules in different orientations. Even frozen samples, however, cannot tolerate an unlimited radiation dose. Individual images therefore contain only weak signals and considerable noise. The methodological advance lies not merely in lower temperatures, but in the interplay between sample preparation, electron optics, detection and computational reconstruction. Only this interplay turns numerous limited views into a three-dimensional description of the molecule or molecular complex under investigation that can be analysed.

Many views do not produce a direct photograph

In single-particle analysis, numerous images of similar molecules are compared and grouped according to their orientation and, where applicable, their different states. A three-dimensional density can be reconstructed from these projections. An atomic model subsequently fitted to it is already an interpretation of this density and should not be equated with the original measurement data. Quality can vary considerably within a structure: a rigid core is often more clearly visible than mobile peripheral regions. Preparation can also favour certain states or cause sensitive components to be lost. Rigorous structural work therefore checks data partitioning, local resolution and agreement between model and measurement. The fact that a reconstruction looks detailed overall does not mean that every side chain or every flexible loop has been determined with equal confidence, or that mechanistic conclusions are justified without additional experiments.

Functional experiments make structures more informative

A three-dimensional structure can show where molecules make contact with one another or through which opening a substance might be transported. On its own, however, it does not prove the complete sequence of events in a biological process. Comparisons of different states, biochemical activity measurements and targeted changes to individual building blocks help with this. Cryo-electron microscopy is established as a powerful route to the structure of many biomolecules. Whether a particular density map supports a proposed mechanism, by contrast, must be justified in each case. Results are particularly valuable when structural predictions enable new, independently testable experiments. The methodological development recognised in 2017 thus does not replace the rest of biochemistry. It provides a three-dimensional basis on which competing explanations can be formulated more precisely and distinguished more effectively through further measurements.

Sources & further reading

Written in our own words with AI assistance, based on the sources listed above. The text distinguishes research findings, technical prototypes and established knowledge. This feature is not a breaking-news report.

Keep exploring

All features on Science