Thanks to the cryo-electron microscopy, the possibility to solve molecular structures up to a resolution of the atomic level has now become a real and feasible alternative to crystallography. The cryo-EM shall dramatically change the way to study proteins and their interactions with drug substances

Cryo-electron microscopy (cryo-EM) is a quite young technique with a huge potential to help increase the structural knowledge of 3D complex macromolecules such as proteins; so huge that it has named the “Method of the Year 2015” 1 according to Nature Methods.

The so-called “single particle cryo-EM” allows for the determination of the structure of a single macromolecule up to the atomic level resolution (2-3 Å): a result that was far to be imagined just ten years ago and that may greatly help in the elucidation of interactions with other biological molecules or with drugs substances.

How it works

The revolution in structural biology started in 2012-2013, whit the appearance of the first commercial instruments for cryo-EM. Since then, the number of scientific articles in the literature has exponentially increased.

The technique itself borned in 1968 from the work of David De Rosier e Aaron Klug2, but the real potential of cryo-EM emerged more clearly only after the invention of the “vitreous ice” technique to prepare the analytical samples3. The first single particle 3D structure (of the 70S ribosome) at a resolution of 11.5 Å was published3 in 2000 by Joachim Franck. The resolution limit of 3 Å was overtaken in 2015: the term “resolution”, explained the expert Sriram Subramaniam during a recent Cell’s4 webinar, has a quite distinctive significance in cryo-EM with respect to traditional X ray’s crystallography. The obtained resolution for different regions of the macromolecule under study might be slightly different, and the resolution refers thus to an average value on the entire structure. The cryo-EM structural determination for b-galactosidase at a resolution of 2,2 Å was published5 in 2015 by the group of the US National Cancer Institute directed by Subramaniam; the results represent the first example of the observation of molecular details such as single amino acids and their side chains, water molecules or metal ions unbedded within the protein tertiary structure. The technique allows also to characterise the dynamic effects of the molecular structure.

A main advantage of the cryo-EM method is the possibility to directly use samples in aqueous solution, thus subjected to random movements. Just 2 µl of solution are enough to run the procedure on proteins of 150kDa’s minimal dimensions. The dimensional limit might diminish to 100 kDa in future years, said Bridget Carragher (National Resource for Automated Molecular Microscopy, NRAMM) during the webinar. After deposition on the analytical support and removal of the excess of liquid, the sample is frozen with liquid ethane, giving rise to the “vitreous ice”. The entire process is run automatically by the cryo-EM apparatus and the characteristics of such a material are quite different from those of normal ice crystals.

After insertion in the cryo-electron microscope, a perpendicular beam of electrons passes through the sample and a photographic image is produced (at up to 400 photograms per second), where light and shadow areas represent the geometrical projection of the real particle on the image plane. Bi-dimensional geometrical patterns observed represent all the possible orientation of the molecule within the aqueous solution. Single photograms representing the same projection are then automatically grouped by the software and summed to obtain an higher resolution image thanks to a better signal-to-noise ratio.

To obtain high resolution images it is important to balance the intensity of the electrons beam and the need to avoid damages of the protein structure. Once, for each projection, the optimal view has been acquired, the different views are summed to obtain the tri-dimensional structure. According to what Yifan Cheng, investigator at the Howard Hughes Medical Institute and University of California, San Francisco, said during the webinar, there is need for approx. 300-500 thousands molecules in order to achieve a 3 Å final resolution. Direct detection devices (DDD) are an emerging technology for image acquisition through capture of electrons instead of photons. The entire procedure is automatized and open source softwares (i.e Leginon and Appion) developed by the NRAMM’s experts is available to support the acquisition and elaboration of data6.

Advantages and applications of the method

Compared to classical X rays’ diffraction, cryo-EM allows for structural determinations of proteins that are difficult to crystallise, or proteic complexes or substrates having different conformations. It is also possible to observe the dynamic motions that occurs within the molecule under study. A great advantage with respect to pharmacological application is the possibility to identify with high precision the single amino acids residues and their ligands, as well as hydrogen bonds and intermolecular interactions. The possibility to work on big macromolecules makes cryo-EM competitive with nuclear magnetic resonance techniques, which are particularly useful for small substrates, explained Sriram Subramaniam.

Cryo-EM has a huge potential for the clarification of pathological mechanisms and it could assist the early identification and selection of drug candidates and the validation of pathological targets. Virology, neurosciences, tumors and immunology are the most promising areas of application, according to Sriram Subramaniam

Yifan Cheng’s group published the structural characterisation of the membrane channel TPVR1, activated by capsaicin, heat and anti-inflammatory agents and part of the biochemical pathway at the base of pain. After the publication in 2013 of a structure at 8 Å resolution8, a refinement to 3.4 Å was obtained using the dose fractionation enabled motion correction technique, making possible to exactly identify the residues involved in the pores’ formation.

The bonding site for inhibitors of b-galactosidase, such as PEGT, has been studied by the Subramaniam’s group; the researchers also investigated dynamic enzymes such as the glutamate receptor9, characterised by a very high conformational eterogeneity, and the magnesium ionic channel, resembling a pentagonal geometry10. This is lost when the concentration of magnesium ions is low: the great conformational changes observed disrupt the symmetry of the molecule.

Drug design may also benefit from cryo-EM, as shown for example by the elucidation of the allosteric inhibition mechanism of the dynamic enzyme p97 by a molecule being part of the anti-tumoral pipeline of the National Cancer Insitute11. The structural determination at 2.3 Ă… resolution allowed for the identification of conformational changes of some key side chains, of the binding site for the inhibitor and of conformational changes deriving from the binding of ATP and ADP molecules.

Open issues to be solved

The high costs of the cryo-EM instrumentation is the main entrance barrier which currently limits the wide availability of the technique outside the still few specialised labs worldwide. The creation of central service providers appears to be the approach of choice for many countries, in a similar way to what occurred in the past for particle’s accelerators. The UK’s Bio-Imaging Centre (eBIC) and the Dutch’s Centre for Electron Nanoscopy (NeCEN) are some examples of such approach within Europe; we report in the box about the first cryo-EM to be soon installed in Italy. The NRAMM and the Janelia Research Campus founded in 2006 by the Howard Hughes Medical Institute are among the main centres of cryo-EM in the U.S.

There are still no guidelines or best practices available to help the work of researchers in this emerging field of structural analysis. The publication of raw data should also reach a standardised consensus, suggests Michael Eisenstein in an article published in Nature Methods4.

The first cryo-electron microscope in Italy

Martino Bolognesi

The University “Statale” of Milan announced in June 2016 the acquisition of the first single-particle cryo-electron microscope in Italy. The initiative represents a very important step forward for the entire Italian scientific community in order to maintain a competitive position within the global structural biology framework.

Martino Bolognesi is the head of the Structural Biology Department at the Milan’s University that will host the new instrument. «We started to discuss this possibility in fall 2014 – he tells to Pharma World. – We realised an historic breakthrough was going on in electron microscopy; it is something converging with the crystallographic studies of proteins and macromolecules we were running. The University’s awareness for scientific progress and the funding we received from the Fondazione Invernizzi allowed to start the procedure and we expect the new instrument to be installed by early 2017». The cryo-EM apparatus will also be able to run electron cryo-tomography analysis, which applies to bigger molecular complexes with respect to cryo-EM as well as to entire cells.

The operation has a global value of € 3.5 million and operative costs estimated in € 100 thousands per year. The new cryo-electron microscope will be installed in a special building optimised to reduce vibrations and for the containment of magnetic fields. The new centre will operate as a service provider for all the researchers of the University of Milan as well as for other scientists of the Italian structural biology community. «When we started the operation, the made contacts with other universities in the area and research centres, i.e. San Raffaele Hospital, Humanitas, the National Institute of Molecular Genetics and the National Council of Research. They all showed a marked interest in our initiative», further tells Bolognesi. A specific Regulation will be issued by the University of Milan to establish rules to access the instrument. The service could, for example, be provided as a simple measurement of the sample, giving rise to raw data that should then be directly analysed by the user. «A more intensive service might include also the analysis of the acquired images. Our interest is that the instrument will work H24, seven days a week. We will run a deep analysis of administrative and operative costs in order to establish the tariffs for the service», explains the director of the Structural Biology Dept.

Professor Bolognesi is also involved in the selection of the highly specialised human resources needed to run the new cryo-EM instrument; the competences are quite new and they might come both internally from the University of Milan as well as from a newly open position of associate professor for which the selection of candidates in on going.

The researchers of the Structural Biology Dept. are mainly involved in the study of macromolecular and multi-protein complexes with molecular weights higher than 100 kDa. «Our research with the new instrument will focus on the development of new drugs. Proteins and membrane’s complexes, for example, are the target of 30-40% of the most active medicinal products currently on the market», tells Bolognesi. This sort of biological macromolecules are very difficult to crystallise and they might be investigated in relation to the action of modulator agents, drugs or cellular stress conditions. Viral proteins might also represent the target for new viral drugs. «The resolution potential of single particle cryo-EM is the election method for this type of projects. It is also possible to investigate systems not requiring an atomic resolution, such as lipidic vescicles and subcellular particles or transcription complexes», further explains the expert.


  1. Nature Methods 13, 1 (2016) doi:10.1038/nmeth.3730
  2. Nature Methods 13, 19–22 (2016), doi:10.1038/nmeth.3698
  3. Biophys. J., 110, 4, 756–757 (2016), doi:
  4. Cell, 100, 5, 537-549 (2000) doi:
  5. Cell,
  6. Science, 348, 1147-1151 (2015), 10.1126/science.aab1576
  8. Nature 504, 113–118 (2013) 10.1038/nature12823
  9. Nature, 514, 328-334 (2014),
  10. Cell, 164, 747-756 (2016), doi:
  11. Science, 351, 871-875 (2016), 10.1126/science.aad7974