Combining cryo-electron microscopy and mass spec-based protein analysis, researchers from the Swiss Federal Institute of Technology in Zurich have determined the structure of the 39S large subunit of the mammalian mitochondrial ribosome.
Detailed in a paper published last month in Nature, the effort offers new insights into the function of this subunit as well as a demonstration of the power of EM and mass spec to generate structural information on complexes not amenable to elucidation via conventional crystallography, Ruedi Aebersold, an ETH Zurich researcher and author on the paper, told ProteoMonitor.
The work "shows that really important and quite large structures can be approached by this kind of hybrid technique with a really good outcome," Aebersold said, adding that the 4.9 angstrom resolution the researchers achieved "is very good even by crystallography standards."
As the Nature authors noted, mitochondrial ribosomes are responsible for synthesis of mitochondrial proteins, including the highly hydrophobic mitochondrial membrane proteins involved in aerobic respiration. The 39S subunit is involved in the synthesis of these proteins and their insertion into the mitochondrial membrane.
"It has been a long-standing question how these very hydrophobic proteins can be produced and inserted into the mitochondrial membrane," Aebersold said. "And this structure points towards ways that this might happen."
Essentially, he said, the emerging proteins appear to exit the ribosome through "a sort of docking platform" that allows them to be directly inserted into the mitochondrial membrane. The EM-mass spec-based structure also identified significant changes in mitochondrial ribosomes compared to their bacterial counterparts, including the replacement of rRNA by proteins and remodeling of the unit's tunnel exit region.
The researchers used cyro-EM to essentially determine the shape of the complex," Aebersold said. "It shows the protrusions and basically the envelope of the complex at good resolution."
EM alone isn't sufficient, however, "to say which proteins are at what protrusion and how they all fit together," he said.
For this information, the researchers turned to a mass spec workflow in which they first chemically crosslinked the proteins of interest then followed with analysis on a Thermo Fisher Scientific Orbitrap Elite instrument.
By crosslinking the proteins prior to mass spec analysis, the researchers were able to determine not only what proteins are present in the complex, but their proximity to one another as well.
"If we know we have, for instance, six proteins in a complex and we know the crosslinks between each one of those, then we can orient them and say that this is how they are arranged," Aebersold said. "And then if you have the structural information from EM or another source, you can use the orientation and neighbor relationships to go towards the structure."
The notion of obtaining structural information via mass spec analysis of crosslinked peptides is not a new one, Aebersold said, noting that his work in this area emerged from conversations with University of Washington researcher David Baker, an expert in computational modeling of protein folding.
Baker "always said that if he had a few [experimentally determined] constraints to tell him which [amino acid] residues were in close proximity that it would be very helpful for the modeling of protein structures," Aebersold said.
Since then, the approach has evolved to become "a very robust technique," he said, noting that the 39S subunit is one of the larger structures solved to date using the combined EM-MS method.
Advances in the technique are due primarily to improvements in mass spec instrumentation and the informatics underlying the identification of crosslinked peptides, Aebersold said.
Spectra generated by mass spec analysis of crosslinked peptides are more complicated than those of unlinked peptides, and therefore require additional informatic capabilities beyond those used for conventional proteomic analyses.
"So we have created software that allows us to assign peptide sequences to these crosslinked peptides, and it now works quite well," Aebersold said.
He added that advances in integrating outside data from sources like EM or small angle x-ray scattering have also helped improve the method.
The technique is not intended as an improvement over conventional crystallography, Aebersold said, noting that "a good crystal structure will have a higher resolution." But, he said, it provides "another approach to the structural biology arsenal in case where a crystal structure is not obtainable."
In the case of the 39S subunit, crystallography was challenging due to the difficulty of obtaining large enough amounts of mitochondrial ribosomes.
"This is a large structure that is basically in an organelle in a cell, and so it has simply not been possible so far to isolate enough of it to do crystallography," he said. Sample requirements for the EM-MS approach, on the other hand, are orders of magnitude less than what's needed for crystallography.
"That's why we think this is an interesting technique," he said. "It will allow people to get good resolution structures of very complicated molecules from very small amounts of material."
Aebersold said that he is currently collaborating with ETH Zurich researcher Nenad Ban, co-author on the Nature paper, to use the technique to explore other aspects of ribosome biology, including the binding of various factors that help control the complex's protein machinery.
His lab is also involved in several related projects looking at chromosome remodeling complexes, he said.