Cutting through complexity


Rebecca Pool

Thursday, June 27, 2019 - 14:30
Image: Professor Wolfgang Baumeister, has determined the molecular architectures of some of the world's most intricate biological structures. [Jung Foundation for Science and Research]
While Professor Wolfgang Baumeister's contribution to the world of cell biology has been nothing short of profound, his performance at high school was, in his words, 'not overwhelming'.
Now famed for his work on the protein complex, proteasome, which degrades unwanted proteins or proteins no longer needed within cells, everyday school lessons did not inspire.
However, a three-year project on plant ecology sparked a life-long interest in biology.
“I was looking at the interactions within plant communities, which was something I could do without any access to fancy instrumentation,” he says. “It was based on observation and description, and to some extent a little mathematical analysis of the data.”
Awards for his work ensued and on leaving school, Baumeister swiftly moved to the University of Münster, Germany, to study a general science diploma.
One year in, he was drafted into the military, but after a brief stint with the German Air Force, returned to academic life at the University of Bonn. As he jokes: “I was not very successful in defending my country.”
Early tomogram of lipid vesicles [Baumeister et al]
Settling into the Department of Agricultural Botany at Bonn, Baumeister quickly took on a cell biology project looking at nitrogen fixing symbiotic association.
Crucially, the department had an unused Zeiss electron microscope, so the young student set it up and swiftly completed his project.
“By today's standards the microscope was extremely simple and primitive, but this was such fun and I had to do everything including aligning the microscope, working in the darkroom and teaching myself sample preparation,” says Baumeister. “My supervisor had never touched an electron microscope so just gave me complete freedom.”
One year later, with diploma degree in hand and wanting some professional training, Baumeister joined the Institute of Biophysics and Electron Microscopy, University of Düsseldorf. It was 1970, and Baumeister had, in his words, 'struck lucky', and got an early research associate position.
What's more, his mentor was the very busy Helmut Ruska, Director of the Institute and brother of electron microscope pioneer and Nobel Laureate, Ernst Ruska.
An early pioneer  of biological electron microscopy, Helmut Ruska had imaged viruses, bacteria and blood cells, and was now keen to see if heavy atoms could be used to label the molecular constituents of membranes.
The project wasn't easy but Baumeister eventually obtained single-atom images without any biological context using an extremely radiation-resistant organometallic compound of thorium.
Still, as he points out: “My supervision was very relaxed essentially being asked once a year how I was getting one, and I totally enjoyed this.”
Professor Wolfgang Baumeister: “I will not stop doing science, and I am certain that in-situ cryo-tomography will become one of the dominant methods in structural biology... I see no reason why we cannot approach near-atomic resolution." [Jung Foundation for Science and Research]
Baumeister obtained his PhD from Düsseldorf in 1973, and continued research here for the next eight years, initially hoping to solve electron microscopy's key problem of radiation damage.
“At the time there was a lot of scepticism amongst biologists that TEM would produce anything close to my dream of near-atomic resolution, so I wanted to find a way to combat this radiation damage,” he says.
He experimented with the use of radical scavengers and other methods to alleviate damage and also turned to spectroscopy to study radiation effects, but come the late 1970s, realised his work was in vain.
Indeed, researchers from the Medical Research Council Laboratory of Molecular Biology, Cambridge, UK, Biozentrum of the University of Basel, Switzerland, and UC Berkeley Lab, US, were developing successful strategies based on low electron dose imaging and averaging.
And so Baumeister changed direction, and focused on revealing the structure of the surface layer protein of a bacterium with legendary radiation resistance, the Deinococcus radiodurans.
Convinced that the future of electron microscopy, hinged on image processing, Baumeister joined forces with Olaf Kübler from ETH Zurich, who was developing computer processing methods for electron microscope images of biological structures.
Progress was rapid and they soon resolved the molecular architecture of the D. radiodurans cell envelope.
Averaging methods
At around the same time, Baumeister also became interested in emerging methods for averaging single molecules as a means to improve resolution with imperfect 2D crystals. 
He initially collaborated with recent Nobel Laureate, Joachim Frank, then at the NY State Department of Health at Albany.
But after failing to fully apply correlation-based averaging to micrographs of the hexagonally-packed-intermediate-layer from the cell envelope of the D. radiodurans, he moved to the Cavendish Laboratory, University of Cambridge. Here Baumeister joined forces with Owen Saxton, who was developing powerful image processing software.
Results ensued quickly and by 1982 - within only a year - the pair had adapted correlation averaging to an almost fully-automatically image averaging method to determine near-periodic structures, even with crystal imperfections.
They applied the method to electron micrographs of the same HPI‐layer from D. radiodurans, resolving the structure to 1.9 nm and publishing results in the Journal of Microscopy. 
As Baumeister points out: “This has had so many citations, not so much for the main methods for correct lattice distortion, but more for its introduction of a new resolution criterion, the Fourier shell correlation.”
“This was such a smart idea and is still used today, especially in single particle fields,” he adds.
Herpes simplex virus [Baumeister et al]
Shortly afterwards, Baumeister moved to the Max-Planck Institute of Biochemistry, Martinsried, Germany, where he was appointed Group Leader. With plentiful resources now at his disposal, work on 2D arrays continued apace.
At the same time, Baumeister and his group also extended these structural studies to more bacterial surface layers and adopted new analysis methods.
“We suddenly had the means to do protein sequencing, DNA sequencing and more,” he says. “We could take a much more comprehensive approach... and with the extra stability we also took on riskier projects.”
It was the late-eighties, and the 20S proteasome from eukaryotic cells was relatively uncharted territory. While other groups had isolated and characterised the protein complex, its active sites and enzymatic mechanism remained elusive.
Given this, Baumeister and colleagues sought out simpler proteasomes in prokaryotic cells, and found the protein complex in the microbe, archaeon Thermoplasma acidophilum. 
By 1991, they had produced the first low-resolution 3D structure of the complex with a more definitive 3D structural model following in 1992 that stood the test of time.
The group also soon expressed the fully assembled and functional 20S proteasome in Escherichia coli, which proved critical for mutagenesis studies and advances in crystal growth. And using electron microscopy with image analysis, they provided the first description of the 26S proteasome; the 20S complex is now known to be a part of this massive complex.
26S proteasome [Baumeister et al]
Come the mid 1990s Baumeister was using X-ray analysis with Nobel laureate, Robert Huber, to resolve the crystal structure of the 20S proteasome.
Using isomorphous replacement and cyclic averaging, success ensued with detail down to 3.4 angstrom resolution published in Science, in 1995.
Other complexes followed, but by 2000, Baumeister had also fully sequenced the genome of Thermoplasma acidophilum. This truly kick-started his all-important use of cryoelectron tomography to map the cellular proteome of the archaeon.
“There was enormous scepticism and people were making jokes about looking at whole cells but because of the steady and stable financial support we had [at Martinsried], as well as the complete freedom, we could pursue this avenue of research,” says Baumeister. “Initially we looked at simple lipid vesicles and other structures and it was clear to us that cryoelectron tomography had enormous potential for structural cell biology.”
Tomography and beyond
Looking back, Baumeister points to automation as being the all-important development that has helped him to drive cryo-electron tomography forward, but the tricky task of collecting data over a wide tilt range with small-angle increments, and with a low electron dose, stymied progress.
The advent of computer-controlled electron microscopes and large-area charge coupled device (CCD) cameras in the late 1980s changed this, with Baumeister quick to seize the opportunity to automate tomographic data acquisition.
“Automation changed everything as we could take a whole tilt series with an acceptable electron dose,” he says. “At this time, microscopes were still operated mostly manually so we had to build the hardware for automation... but CCD cameras also allowed us to establish a feedback loop and automate data acquisition.”
By the late 1990s, Baumeister and colleagues were already using automated electron tomography to study lipid vesicles, actin networks and cytoplasmic complexes in cells embedded in ice.
Importantly, they augmented their studies with the development of image analysis software, such as 'template matching' with the 'TOM' software package which streamlines all cryo-electron tomography processing steps following later.
With automation and software in hand, the researchers soon resolved the 3D structure of the Herpes simplex virus using cryo-electron tomography with averaging to improve the signal-to-noise ratio of the tomograms.
And incredible images of the nuclear pore complex within the nuclei of the Dictyostelium discoideum amoeba followed. 
But while image processing brought more and more cryo-electron tomography results, specimen thickness posed a problem.
Cryo-sectioning methods introduced artefacts, so Baumeister and colleagues quickly turned to an alternative slicing method; focused ion beam technology.
With a grant from the Germany-based Federal Ministry of Education and Research they converted a dual-beam focused ion beam/SEM instrument into a cryoFIB by adding a cryostage and adapting the instrument for cryoSEM preparation. They then went on to use the method to image 200 nm lamellae from vitrified eukaryotic cells.
Cryo-electron tomography reveals the HeLa cell nuclear periphery [J. Mahamid, Baumeister et al]
Baumeister highlights how cryo-FIB development continues, but only recently his group used the method to obtain incredible 3D snapshots of previously elusive structures in-situ at the periphery of the nuclei of HeLa cells.
“Focused ion beam micromachining really makes any volume of the cell accessible,” explains Baumeister. “We can now cut windows into the cell and this has opened up the field of cellular structural biology, and what we now like to call structural biology in-situ. But mining the rich information of tomograms remains a challenge.”
Baumeister has always emphasised the importance of coupling his methods with correlative techniques to, for example, identify and target regions of interest within samples.
As early as 2007, he and colleagues were using cryo-electron tomography with light microscopy to image neurons grown in culture. And more recently the researchers have used cryo-FIB with correlative light and electron microscopy to prepare neuronal cells for cryo-TEM analysis of toxic protein aggregates.
“We discovered that a key player in aggregate interactions is the 26S proteasome so I think we have closed a circle here,” says Baumeister. “It really is gratifying to see how our biology-driven projects, enabled by our technology developments, converge.”
Neurotoxic aggregate populated with 26S proteasome [Baumeister et al]
Cryotomography-related developments will no doubt continue. Only a few years ago, Radostin Danev in the Baumeister lab developed the Volta phase plate for phase contrast cryo-EM.
This phase plate is designed to replace the Zernicke thin film phase plate, which is short-lived, difficult to use in automated data acquisition routines and can introduce fringe artefacts in images.
Looking forward, Baumeister now intends to focus on improving software for extracting more information from tomograms.
Retirement from the Max-Planck Institute of Biochemistry is on the horizon, but the researcher is already setting to up a new group at Shanghai Tech University, China, which will be up and running by the end of 2021.
Baumeister's latest plans are quite literally a world away from his early Düsseldorf research into radiation damage to reach near-atomic resolution. However, his resolve remains.
“I will not stop doing science, and I am certain that in-situ cryo-tomography will become one of the dominant methods in structural biology,” he says. "I hope that in the next three years we will routinely reach 5 angstroms resolution, and as small improvements accumulate I see no reason why we cannot approach near-atomic resolution.“
[Thanks to Jung Foundation for Science and Research for all photos of Professor Wolfgang Baumeister.]
Website developed by S8080 Digital Media