The Resolution Revolutionary


Rebecca Pool

Thursday, October 4, 2018 - 15:45
Interest in an obscure branch of physics and faith in electron microscopy led to Professor Richard Henderson winning his Nobel Prize. Rebecca Pool reports.
Late last year, Richard Henderson of the Medical Research Council Laboratory of Molecular Biology, (MRC LMB) in Cambridge, UK, won the 2017 Nobel Prize in Chemistry alongside fellow electron cryo-microscopy pioneers Joachim Frank, Colombia University, US, and Jacques Dubochet, University of Lausanne, Switzerland.
At a time when researchers worldwide are rapidly unmasking structure after structure of biology's most intricate molecules, the award for the technology that makes this happen was not a big surprise.
Indeed, as Henderson tells Microscopy and Analysis, he and colleagues had long wondered if the Nobel Committee would recognise their work or opt for the much-coveted CRISPR-CAS9 genome editing tool. "But here we are now with a Nobel Prize and I don't think anyone is surprised really," he says.
Henderson's comments sound low-key, but the impact his research has had on structural biology cannot be understated.
The structure of bacteriorhodopsin at 7 ångström in 3D, from 18 images and 15 diffraction patterns. Top left: late electron diffraction pattern, top right: 1975 balsawood model of bacteriorhodopsin molecule [Henderson and Unwin, Nature, 1975, 257, 28-32], bottom: freeze fracture image [Walther Stoeckenius]
Frustrated with the unwieldy sample crystallisation necessary to characterise biomolecules using X-ray crystallography, he and colleague Nigel Unwin, smothered his bacterial cell membrane containing the bacteriorhodopsin protein with glucose and placed it in an electron microscope.
The first and finest electron microscopy 3D images of the protein followed, it was only 1975, and Henderson had made the first tentative steps to kick-start a structural biology revolution.
Grass roots
Schooled in Edinburgh, Mathematics and Science came 'fairly naturally' to the young Henderson, while other subjects didn't. "I tried to get rid of all the languages but was then told by my headmaster that if I didn't study Latin or French I wouldn't be able to go to university," he says.
Henderson took his headmaster's advice, studied the O-Level equivalent of French and was the only one in his class to fail the exam.
"Fortunately another girl had failed her Higher exam, so the two of us spent an entire year learning French with one teacher," recalls the Nobel Prize winner. "I scraped a pass but at least knew I was then allowed to go to university."
At around the same time, an enthusiastic Physics teacher had also insisted Henderson's class listen to an entire 40 minute lecture on a little known subject, Biophysics.
"I don't know how much we got out of the talk other than learning that our teacher thought Physics and its impact on Biology was going to be very important," laughs Henderson. "But he was very enthusiastic and out of a class of around twenty, four or five of us signed up to study Physics at Edinburgh University."
At college, Henderson tried to eliminate all subjects except for the ones he was really interested in. His course consisted of Physics, Mathematics and Mathematical Physics, and as he puts it: "I have always thought that if you spread yourself too thinly across many subjects you can't get very deeply into any of them."
Meanwhile, he also worked in the Physics department as well as taking on other jobs in laundries, delivery rounds and more.
"My father had been a baker all of his life and many of our friends had retired in the same job they started in," he says. "But I just wanted to figure out what was going on in the wider world, and there wasn't a whole lot of career advice out there."
Curiosity and determination led to academic success, and come the end of his degree, Henderson knew he wanted to study Biophysics. "I didn't want to be part of a vast team of 1000 people studying, say, gravitational waves," he says. "I was interested in doing things myself without being told what to do, and biophysics fitted that."
Still, it was the mid-1960s, and amid the deluge of detail on plasma physics, high energy particle physics, solid state physics and even astrophysics, the young student struggled to unearth biophysics information. Yet, after a few false starts, he eventually accepted a PhD offer from the Biophysics Department at King's College London and prepared to move to the city. And then he met Professor Bill Cochran.
Cochran had joined Physics at Edinburgh after working in Cambridge where he knew about the MRC LMB, and as Henderson highlights: "He said no, no, no to King's, knew everyone at Cambridge including Max Perutz, so I wrote to Perutz who immediately invited me to visit."
The young student loved the MRC LMB instantly. "It was an open day on a Saturday but everyone was working much harder and seemed more interested and cleverer than anywhere else I had visited at the time," he says. "I have always believed it is so important to leave no stones unturned and find out everything that is available... and after months of reading, talking to people and visiting places, I had found this incredible laboratory in Cambridge."
A slice through a 3.5 ångström resolution 3D map of bacteriorhodopsin with the corresponding atomic model superimposed. [Henderson et al, J Mol Biol, 1990, 213, 899-929
Perutz, who had shared the 1962 Nobel Prize for Chemistry with John Kendrew for studies of the structures of haemoglobin and myoglobin, offered Henderson a three-year Scholarship to look at the proteolytic enzyme, chymotrypsin. Within months, he was well trained in X-ray crystallography, had helped to resolve the structure of this protein-hydrolysing enzyme, publishing results in Nature.
Come 1970, and with PhD in hand, Henderson crossed the Atlantic to take up a postdoctoral position at Yale University focusing on the structure of bacteriorhodopsin, a purple-coloured protein embedded in the membrane of a photosynthesising bacterium. But progress using X-ray diffraction was slow as coaxing these membrane-embedded proteins to crystallise was proving fruitless.
Three years later, Henderson returned to MRC LMB, and everything changed. He swiftly encountered Nigel Unwin, who had been using electron microscopy to image the tobacco mosaic virus. Henderson began to wonder if he could use electron microscopy to record images and electron diffraction patterns from bacteriorhodopsin, without negative stain.
For the next two years, the pair worked avidly on this proposition, with Unwin switching from viruses to bacteriorhodopsin and Henderson abandoning X-ray crystallography for electron microscopy. This time, rather than trying to remove the protein from the membrane, Henderson and Unwin took the complete membrane, embedded it in a thin film of glucose, and placed it directly in the electron microscope at room temperature.
The carbon, nitrogen and oxygen atoms of biological structures only gave low-contrast pictures, but the researchers also obtained electron diffraction patterns from which they could calculate a more detailed image. By tilting the sample and taking images through many angles, they went on to resolve 3D structure of the membrane protein at a resolution of 7 Ångströms. In 1975, this was the best-ever image of a protein from an electron microscope.
'Even now, there are so many developments being made and these are going to make electron cryo-microscopy a really, really fantastic method.' Richard Henderson
"Our little collaboration had proven to be quite successful and we continued for decades working like this," says Henderson. "For example, we spent well over a year and a half trying to find out where the retinal that gives bacteriorhodopsin its purple colour was located."
"We tried labelling samples with heavy atoms or fluorescent probes without success, but eventually got good results with neutron diffraction," he adds. "We were trying many, many different things and it all took a long time actually; some of our ideas worked and some of them really didn't."
Henderson continued to refine and add more details to his model of bacteriorhodopsin, but by the early 1980s was convinced that electron cryo-microscopy could produce high quality images and wanted to pursue imaging rather than diffraction.
"We didn't have the microscopes and cold stages at Cambridge at the time; we could get images, but vibrations, drift and technical problems with hardware meant we couldn't get images that were better than 7 Ångströms," he says. "We'd already got the structure to this resolution in 1975 without having to cool the sample, so what was the point?"
And so Henderson took trips to several laboratories that housed home-made electron cryo-microscopes. First stop was a hybrid Zeiss-Siemens microscope with a superconducting liquid-helium objective lens, built by now fellow-Nobel Prize winner, Jacques Dubochet, in the European Molecular Laboratory, Heidelberg.
Progress was difficult but an image showing diffraction beyond 4 Ångströms was produced, prompting Henderson to develop computer processing methods for high resolution 2D crystals of unstained membrane proteins.
Next, Henderson visited the Fritz-Haber Institute of the Max-Planck Society in Berlin, sparking a decade-long collaboration with Fritz Zemlin and Erich Beckmann. By 1986, using a Siemens 100 keV electron microscope with a superconducting lens and conventional tungsten source, the researchers had obtained high resolution images of bacteriorhodopsin in projection.
Henderson and fellow electron cryo-microscopy pioneer, Ken Downing from Berkeley Laboratory, also recorded some cryo-EM images from purple membranes - showing diffraction beyond 4 Ångströms - on Downing's JEOL 100B.
Henderson persevered with processing these high resolution images, but was also working on microscope hardware. Key improvements included the correction of beam-tilt misalignment, developing a method to correct for the gradient of defocus from differently-tilted samples, and using spotscan imaging to reduce beam-induced image blurring on highly tilted specimens.
"In Cambridge, we had also decided that all the cryo-EM cold stages were really bad so we built out own in our workshop, eventually licensing the technology to Oxford Instruments, later bought by Gatan," points out Henderson. "I think this was an area where you really had to do some work yourself, but you also had to work with and persuade the commercial companies to improve [your developments]."
The resolution of electron microscopes has radically improved in the last few years, from mostly showing shapeless ‘blobs’ to now being able to visualise proteins at near-atomic resolution. [Martin Högbom/The Royal Swedish Academy of Sciences]
Finally, in 1990, still recording images on photographic film, and fifteen years after his 7 Ångströms model, Henderson determined the structure of bacteriorhodopsin to 3.5 Ångströms, near-atomic resolution. He had finally proven that electron cryo-microscopy could generate astonishingly detailed images of biomolecules, just like X-ray crystallography.
"In the end, probably by 1993 you were beginning to get commercial cryo-EMs that were vaguely usable," says Henderson. "And now we have had lots of technical developments in the vacuums, cold stages, automation, field emission guns, and of course CMOS direct electron detectors, making the method what it is now."
The big picture
Henderson's vision of using electron microscopy to image structures atom by atom was a critical driving force behind the rise of electron cryo-microscopy.
At the same time, Joachim Frank from the New York State Department of Health, US, had been developing image processing methods for 3D structures that sharpened 2D images and allowed these to be combined into a 3D structure. And Dubochet had created a way to flash-freeze samples with liquid ethane and liquid nitrogen yielding incredibly well-preserved specimens.
By combining all of these breakthroughs, electron cryo-microscopy resolutions were pushed further and further down. But as Henderson says: "My own view, and I have been saying this for a number of years, is that Jacques Dubochet has been the key person here."
"The plunge-freezing method that his group developed is what really made electron cryo-microscopy what it is, and as soon as we all knew about this, we realised what a very important method it was," he adds. "I would have been personally happy if only Jacques Dubochet had got the Nobel Prize."
But the electron cryo-microscopy resolution revolution doesn't stop at the Nobel Prize. Increasingly detailed images of ever-larger structures - from the Zika virus to complex molecular machines - continue to roll in, and Henderson believes there is much more in store.
For starters, he and colleagues at the Rutherford Appleton Lab, UK, have been working on CMOS detectors since 2002, and hope that a commercial product with unprecedented quantum efficiency will be released in the next two years. What's more, the team have also been looking at using detectors with electron microscopes at lower operating voltages.
"We think that for structural biology, it might be better to [operate the electron microscope] at 100 kV instead of 300 kV," points out Henderson. "Right now there is not a very good detector at the moment for doing this."
Henderson is also excited about innovations in specimen support, pointing to research from MRC LMB's Chris Russo. Right now, samples typically reside on carbon foils during imaging, but the ex-Harvard physicist has developed all-gold supports that reduce radiation-induced motion and image blurring during electron cryo-microscopy.
"People are also developing derivatised graphene supports, so these and gold supports will both increase resolution," says Henderson.
And finally, the Nobel Prize winner is also looking forward to the eventual use of liquid helium, instead of liquid nitrogen, in electron cryo-microscopy.
As he highlights, researchers in Japan, especially Yoshinori Fujiyoshi, have already shown that a further reduction of specimen temperature from liquid nitrogen to liquid helium improves the cryo-protection factor of specimens by at least two-fold.
"There's a whole load of technical issues that need to be solved with liquid helium, but once solved we're going to see some very big improvements," he highlights. "You see, even now there are so many developments being made and these are going to make electron cryo-microscopy a really, really fantastic method."
Website developed by S8080 Digital Media