Hitting new highs

Macromolecular crystallography in 2016

tl_files/root/news/0_Success_Stories_2017/dectris_share_PDB.png
Figure 1. Percentage of X-ray structures in the PDB determined with data from DECTRIS detectors.

Last year, the Protein Data Bank (PDB) released 10898 structures, a new record. Of these, 10012 were determined by X-ray crystallography. Structures of membrane proteins continue to account for a large number of high-profile publications (e.g. Mansoor et al., Zheng et al., Oswald et al., Celia et al., Morales-Perez et al.). Other highlights last year include Zika and CRISPR proteins (Barba-Spaeth et al., Lei et al., Dong et al., Yamano et al., Yang et al.). No fewer than eight high-profile publications investigate regulated proteolysis (Martinez-Zapien et al., Lechtenberg et al., Morgan et al., Cavadini et al., Petzold et al., Schrader et al., Trentini et al., Bhogaraju et al.), making it the hottest topic in structural biology. We would like to congratulate all involved researchers on these impressive achievements.

Structural biologists increasingly rely on DECTRIS products for their work. Last year, 30% of all released PDB entries were determined with PILATUS or EIGER data (Figure 1). You use our detectors because you will get data of the highest possible quality, which allows you to tackle the most challenging projects. High-profile publications are the deserved reward. In 2016, data from DECTRIS detectors contributed to 54% of Cell, Science and Nature papers that reported crystal structures.

In 2016, the first two fourth-generation synchrotrons [1], MAX IV in Lund, Sweden, and NSLS-II on Long Island, United States, reported first light. Many other synchrotrons prepare major upgrades that promise a dramatic change in how macromolecular crystallography is done. New detectors are an important part of all upgrade plans. EIGER has already arrived at MAX IV and NSLS-II and a number of other facilities (Table 1).

SynchrotronBeamlineDetector
ASLS PX2 EIGER X 16M
APS GM/CA EIGER X 16M
APS LS-CAT EIGER X 9M
APS NE-CAT EIGER X 16M
ESRF MASSIF-3 EIGER X 4M
MAX IV BioMAX EIGER X 16M
NSLS-II AMX EIGER X 9M
NSLS-II FMX EIGER X 16M
Photon Factory BL-1A 2x EIGER X 4M
SLS X06SA EIGER X 16M
SOLEIL PROXIMA-2A EIGER X 9M
SPring-8 BL26B1 EIGER X 4M
SPring-8 BL32XU EIGER X 16M
SSRL 12-1 EIGER X 16M
Table 1. EIGER detectors are installed at 14 macromolecular crystallography beamlines worldwide.

Crystallographers who have worked with EIGER (e.g. Joshua Carter of Shamrock Structures and Dan Rosenbaum of the University of Texas Medical Center) have seen improvements in the signal-to-noise ratio of high-resolution reflections. These empirical observations hold up to scientific scrutiny. A paper published by Meitian Wang's group at the Swiss Light Source shows that EIGER measures data with even higher accuracy than PILATUS. This higher quality is already reflected in the PDB. EIGER data are associated with 0.3% of the structures released last year. At the same time, four of the 139 Cell, Science and Nature papers relying on MX data reported data collection with EIGER. At nearly 3% of the total, this is a tenfold overrepresentation.

These are exciting times to be a structural biologist. Cryo-electron microscopy and tomography and electron diffraction are increasingly powerful. Techniques like small-angle X-ray scattering and nuclear magnetic resonance spectroscopy have been developed to give results of unprecedented detail. But it is X-ray crystallography that provides the highest-resolution information for the widest range of macromolecules. We are looking forward to reading how you capitalize on PILATUS and EIGER for more great structures in 2017.

[1] D. Castelvecchi. (2015). Next-generation X-ray source fires up. Nature 525: 15-6

Electrons are the new photons

EIGER tested for the first time in an electron microscope

tl_files/root/news/0_News_2017/dqe_with_energy_40_120kV.jpg
Figure 1. Detective quantum efficiency of EIGER X 1M. The DQE of EIGER X 1M is shown for electron energies between 40 and 120 keV.

Hybrid photon-counting detectors have caused a major breakthrough in macromolecular crystallography. The new EIGER detectors for synchrotron applications feature a pixel size of 75 µm and continuous read-out at frame rates of up to 3000 Hz. In its first year in regular operation at MX beamlines, EIGER has already contributed to several outstanding publications.

Since EIGER was designed for noise-free single-event counting, Wasi Faruqi and Greg McMullan from the detector group at the MRC Laboratory for Molecular Biology in Cambridge (U.K.) wondered how it would perform with electrons instead of X-rays. Better performing detectors are considered a key to further improvements in the resolution of electron cryo-microscopy [1]. To find out how EIGER compares, we collaborated with Faruqi and McMullan to characterize an EIGER X 1M in a Philips CM200-FEG transmission electron microscope installed in the EM suite of the LMB. For all tests, the detector was operated at a frame rate of 2200 Hz and with its threshold tuned to between 5 and 50 keV. The microscope was set to electron energies between 40 and 200 keV.

Using the knife-edge method [2], we projected the shadow of a thin gold wire onto the surface of a specially calibrated EIGER X 1M detector. From the shadow images, McMullan determined the edge spread function and the modulation transfer function (MTF) of the detector. Processing the data in single-event mode and using noise binning [3], he then calculated the detective quantum efficiency (DQE, see figure) of EIGER. At energies of 80 and 100 keV, the DQE curves are very near the theoretical limit across all spatial frequencies. EIGER turns out to be the ideal counting detector for electrons with energies of up to 100 keV [4].

[1] W. Kühlbrandt. (2014). The resolution revolution. Science 343: 1443-4.

[2] A. Faruqi and R. Henderson. (2007). Electronic detectors for electron microscopy. Curr Opin Struct Biol 17: 549-55.

[3] R. Meyer and A. Kirkland (1998). The effects of electron and photon scattering on signal and noise transfer properties of scintillators in CCD cameras used for electron detection. Ultramicroscopy 75: 23-33

[4] Detective Quantum Efficiency of EIGER X 1M for Electron Microscopy. Manuscript in preparation.

High on EIGER

The structure of the cannabinoid receptor CB1

G protein coupled receptors (GPCRs) are key modulators of human physiology. These membrane proteins relay signals from the environment and cause cellular responses. A large share of marketed drugs acts by binding to GPCRs.

Recently, the range of marketed drugs has grown in certain jurisdictions to include cannabis. While it is often used recreationally, cannabis also has medicinal purposes and its primary active ingredient, Δ9-tetrahydrocannabinol (THC), holds promise as a candidate for pharmaceutical development. It has wide-ranging effects on memory, mood, pain, inflammation, and appetite.

THC binds to the cannabinoid GPCR CB1, the most abundant receptor in the brain. Its physiological ligands are endocannabinoids that modulate neurotransmission. For derivatives of THC to become therapeutic agents, the biology of the endocannabinoid system has to be understood better. Without structural information about cannabinoid receptors, the molecular details of ligand binding are still largely unknown.

At the end of last year, two papers appeared in quick succession that fill this gap. First, a group around Ray Stevens from the Bridge Institute at the University of Southern California, Los Angeles, and the iHuman Institute at ShanghaiTech University published the structure of CB1 in complex with the agonist AM6538 (Hua et al.). Recently, a group led by Dan Rosenbaum from the University of Texas Southwestern Medical Center in Dallas added the structure of CB1 in complex with the inverse agonist taranabant (Shao et al.).

Getting there was not easy. Membrane proteins like CB1 are notoriously difficult to isolate. Both groups worked with CB1 constructs with a stabilizing domain inserted in one of the intracellular loops of the receptor. Crystals were grown in lipidic cubic phase.

The Rosenbaum group collected data at the GM/CA beamline at Advanced Photon Source where effective data collection on microcrystals is possible. The two end stations are equipped with a PILATUS 6M and an EIGER X 16M, respectively. "Both detectors have given us a significant boost in signal at high resolution relative to the previous CCD detectors", says Rosenbaum. He adds, "the EIGER detector that we used for CB1 worked very well, similar to the PILATUS we had previously used but perhaps with a slightly better signal-to-noise for our samples."

Reflecting the difficulty of the project, data from crystals of CB1 in complex with AM6538 were collected at a number of synchrotrons. "We currently get excellent data from SPring-8 consistently", says Stevens. Not only is beamline BL41XU finely tuned with strong support, it also provides a fast PILATUS3 6M detector.

tl_files/root/news/0_Success_Stories_2017/CB1_Fig1.png
Figure 1. The structure of cannabinoid receptor CB1. The structures of CB1 in complex with AM6538 (left, PDB code 5TGZ) and taranabant (right, PDB code 5U09) are shown as cartoon representations. The ligands are shown as yellow stick models. The extracellular side of the membrane (indicated by the light green bar) is up. A cannabis leaf is shown in the center.

The two structures of CB1 dramatically advance our understanding of the interactions of small molecules with the receptor. While the structures show the receptor in its inactive conformation, the binding pocket for THC and endogenous ligands could nevertheless be modeled. This information will aide the development of new candidate drugs with fewer side effects. Until then, if you want to stimulate your research, EIGER is a safer bet than THC.

Diamonds are a PILATUS’ best friends

PILATUS data helps shed light on Earth’s deep mantle regions

Diamonds larger than a thousand carats are rare, valuable and fascinating. It now appears it is not only their size but also their origin that makes massive diamonds extraordinary. A recent publication in Science identifies inclusions in large diamonds as metallic phases and suggests that these diamonds grew from liquid metal in Earth’s mantle (Smith et al.). The research verifies predictions that Earth has highly reducing deep mantle regions. The redox state of the convecting mantle is a key unknown in Earth’s evolutionary history.

tl_files/root/news/0_Success_Stories_2017/nestola.png
Figure 1. Fabrizio Nestola with a large diamond in front of the SuperNova instrument equipped with a PILATUS 200K detector. Image courtesy of Fabrizio Nestola.

A group of researchers around Evan Smith from the Gemological Institute of America characterized inclusions in large diamonds. They used a wide range of methods from Raman spectroscopy over scanning electron microscopy to X-ray diffraction. Co-author Fabrizio Nestola from the Department of Geosciences at the University of Padova contributed the X-ray diffraction analysis, which confirmed the prominence of iron and other metals in the inclusions.

For the X-ray diffraction analysis, Nestola used a SuperNova instrument that was the first of its kind to be equipped with a PILATUS detector. The instrument was purchased when SuperNova instruments were still distributed by Agilent and only CCD detectors were offered with the instruments. However, Nestola and his colleague Ross Angel clearly saw the advantages of HPC detectors and insisted on having a PILATUS detector installed. With this detector, they obtained the results that made it onto the cover of Science. Meanwhile, Agilent’s branch for X-ray diffraction has become part of Rigaku Oxford Diffraction, a company that offers PILATUS and EIGER detectors with all its single-crystal diffractometers.

Bringing it all back home

Debye-Rietveld celebration, Amsterdam 2016

Guzmán Peinado with his prize-winning poster

It was not too long after the Braggs’ findings that Peter Debye set another two milestones that revolutionized X-ray diffraction theory. The first was in 1916, when he constructed a setup for diffraction experiments on bulk materials, which made the leap from fundamental to applied science and gave rise to X-ray powder diffraction (XRPD). The second was Debye’s scattering equation, a single relation that redefined the fundaments of the approach by expanding it to non-crystalline materials. These two contributions were so powerful that it took almost fifty years before the technique was transformed again. In 1966, Hugo Rietveld presented a novel approach to model XRPD patterns, a fast and efficient tool for extracting quantitative and structural information from the data that can be used by anyone… with access to a computer.

A lot has happened since, and this year the 50th and 100th anniversaries of the contributions of these great Dutch scientists were celebrated in their home country. The Dutch Crystallographic Society invited experts to review how these groundbreaking equations and their reinterpretations formed contemporary materials science. To look ahead, a generation of young researchers presented their projects and ideas that will shape the future directions of the technique.

One of these motivated researchers is Guzmán Peinado, whose attendance of the event was supported by DECTRIS.

"Being from Uruguay, it is sometimes hard to reach specific scientific communities. And seeing how science plays out in different environments as well as getting to know people with lots of experience and diverse backgrounds is crucial for high-level research" comments Guzmán.

And Guzmán himself is an interdisciplinary researcher. He is a math teacher, currently working on his PhD both at the Inorganic Chemistry department and CRYSSMAT-LAB from Facultad de Química of the Universidad de la República in Uruguay. His interest in physical properties of micro- and nanocrystalline metal-organic frameworks (MOFs) has led him to many different analytical methods, and XRPD is one of them. So, a Debye-Rietveld celebration in Amsterdam seemed like a good place to be. As a mathematician, he fully appreciates the development and connection between Rietveld’s and Debye’s equations, and as a chemist, he recognizes that these approaches can help tailor structure-property relationships of MOFs. So, not surprisingly, Guzmán found the talk of Reinhardt Neder particularly interesting. "Neder’s talk was really inspiring because of his expertise in lanthanides and nanoparticles. His knowledge and confidence about the Pair Distribution Function method was also very motivating. It made me wonder about shifting my focus fully to characterization of nanosystems", says the winner of the best poster prize.

Although his poster addressed characterization of microcrystalline materials, a tendency towards synthesis of nanosystems is clearly visible. Now, with the fully implemented PDF methods, and the new high energy PILATUS3 CdTe detector, there is noting standing in the way to tailor-made MOFs!

We wish Guzmán all the best in his future work, and lots of PILATUS3 CdTe data!

Collecting data with passion

A feature on serviced crystallography

Image used by permission of Joshua Carter, Shamrock Structures LLC.

On the journey from an intriguing biological question to a structure, the nights at the synchrotron it takes to collect data good enough to find answers are probably the most tiresome times for most. Hours in the dark, in front of a battery of glowing screens, the worst crystals always appearing between 3 and 5 in the morning – collecting best data is thoroughly exhausting.

As a graduate student, going to the synchrotron for the first time, things were novel, exciting and fun. After a few years, the joy has worn off for most, and data collection at the synchrotron became a dreary routine, even dreaded by some. And yet, collecting data is essential, the last experimental step in crystallography. The quality of the structure and the insight into biology are decided by it.

The situation is even starker in the pharmaceutical industry. Specialization drives structural biologists into the role of focused crystallographers whose job might be to collect countless datasets from hundreds of crystals soaked with different chemicals for fragment screening. This can be tedious work, but there are people who get excitement out of this, people who approach each new crystal as if it were their new favorite project. The best of these scientists have founded or work for companies that provide data collection as a service.

Two examples of such companies are Expose GmbH and Shamrock Structures LLC. While Expose is singularly focused on collecting data for paying customers, Shamrock Structures is a broader research services company for whom data collection is just one field of operation. Both have in common that their success depends on the efficiency and reliability with which they can collect data and on the quality of the data they provide.

Expose is based in Switzerland and calls the Swiss Light Source (SLS) its home. The labs of Shamrock Structures are near the Advanced Photon Source (APS) in Illinois, US, and most of their data are collected there. Both synchrotrons are among the most productive in the world and are renowned for the quality of their beamlines.

Joachim Diez, founder and CEO of Expose and still frequently at the beamline to collect data, is enthusiastic about SLS's approach of developing their beamlines in collaboration with industrial users. He is often among the beta testers of new features before major upgrades. "All beamlines at SLS provide the best combination of high-end technology and usability for high throughput projects", he says.

All three macromolecular crystallography beamlines at SLS operate Hybrid Photon Counting detectors. Diez prefers beamlines PXIII, whose mini-hutch design makes it the easiest to work with, and PXII, which provides the best compromise of flux and usability of the data to the customers thanks to the common CBF format that PILATUS writes.

tl_files/root/news/0_Success_Stories_2016/shamrock_LS-CAT.jpg
Crystallographers at work. Shamrock Structures scientists Ardian Wibowo and Joshua Carter at LS-CAT at APS. Image used by permission of Joshua Carter, Shamrock Structures LLC.

Joshua Carter, Director of Crystallographic Data Collection at Shamrock Structures, considers APS, in particular the beamlines of the Life Sciences Collaborative Access Team (LS-CAT), the best place to collect data. The ring reliably provides X-rays of high flux and brilliance, and the beamlines are equipped with the latest technology for industrial-grade data collection, like CATS sample changers and, at beamline 21-ID-D, an EIGER X 9M detector. They also run innovative software, and hardware and software come together smoothly thanks to the hard work of exceptional beamline staff. Carter and his team are so pleased with LS-CAT that Shamrock Structures signed up as an Associate Member earlier this year.

An important factor behind this decision was the detector. All Hybrid Photon Counting detectors like PILATUS and EIGER offer superb signal-to-noise ratios but, says Carter, "I see an improvement of spot-to-background intensities with the EIGER X 9M compared to the PILATUS 6M", adding that, "there is really no comparison to be made between the EIGER and other detectors on the market."

Another feature that sets EIGER apart from PILATUS is the speed of data collection. This is more useful than just for collecting 12-15 datasets per hour. Normally, datasets are collected at 40-80 Hz, but if crystals are particularly radiation-sensitive, the speed of data collection can be increased to 100 Hz or sometimes 200 Hz. "This allows us to tailor the permissible dose better than with beam attenuation", explains Carter. "We get better data quality with increasing the speed of the data collection."

For pharmaceutical companies, the advantages of outsourcing data collection are clear. Their scientists can focus on interpreting structures and advancing drug discovery projects rather burning time collecting data. Equally importantly, the administrative burden and the logistics involved with setting up synchrotron time and the traveling to and from the synchrotron can be eliminated, and employees are more productive if they don't have to recover from nights spent at a high-throughput beamline. Any of these aspects has a clear financial reward.

Fully automated or remote-controlled beamlines exist where users can mail samples and obtain data without having to be present on site. For example, the European Synchrotron Radiation Facility provides MXpress, a mail-in data collection service, while at beamline I04-1 of Diamond Light Source, up to a thousand samples per week can be screened for bound ligands – with automatic identification of hits. Both approaches are extremely valuable for routine crystallography and high-throughput ligand screening.

When samples are of variable quality and difficult decisions need to be made during data collection, service crystallography shines. Shamrock Structures' customers benefit from the expertise and experiences of the crystallographers they work with. After collecting data for five years, with two to three twelve-hour shifts a week, Carter practically lives in reciprocal space. "I have seen more crystals and more diffraction patterns and solved more structures than entire structural biology groups at some of our customers", he says.

Diez tells a similar story. "Serviced crystallography provides the best quality in combination with the best productivity without any sample limitations", he says. Remote operated beamlines come close, but they still need local support and an expert crystallographer to collect the data and, says Diez, "well-trained crystallographers doing service crystallography combine the local support and the scientist at home."

tl_files/root/news/0_Success_Stories_2016/expose_wall.jpg
A wall of work. Dry shippers from a large number of Expose customers are awaiting beam time at SLS. Image used by permission of Joachim Diaz, Expose GmbH.

The continued enthusiasm of Carter and Diez for their work is another plus for their customers. They signed up for data collection and for nights at the synchrotron. Where others might see tedium, they get excited by spots on a screen and the chance to get better data more quickly. They care about their customers' projects and put their vast experiences into every single one of them, fighting for superior data in every case.

Pharmaceutical companies appreciate this attention to detail and the effort put into their projects. As they pay only for the time spent collecting data, they get their results at minimal cost. This is especially true when the data are collected on a fast detector like EIGER. Small companies gain twice by not only obtaining better data than they could measure themselves but also having immediate access to a synchrotron in ways otherwise only possible for the largest players who can buy into a beamline. "Customers of any size get their data collected when they want it collected", says Carter.

Crystallographic service providers like Shamrock Structures and Expose are important players in the drive towards better data. They accelerate the process of drug discovery and help reduce cost. We at DECTRIS are excited about the role we play in this ecosystem, and we are proud of the trust Carter and Diez place in our detectors.

Pi computed to 22.4 trillion digits

DECTRIS smashes old world record for precision of pi

How big is pi? Most high school students will shoot off 3.14 without much thinking, but Peter Trüb, software developer at DECTRIS, has a different answer. According to his calculation, which came to a conclusion last week, pi weighs in at 8.5 TB. Calculated to 22.4 trillion digits, "pi is so long it would fill a library of several million 1000-page books if printed on paper", says Trüb.

There are several reasons for calculating pi to ever increasing precision. Mathematicians would like to find out whether pi is normal. If so, each sequence of digits would be equally likely to occur, and the stream of digits would appear perfectly random. Computer scientists use pi to test numerical analysis algorithms. For DECTRIS, the point of the calculation was the massive stressing of a computer system similar to those under development for EIGER detector control and data processing.

We used a DELL PowerEdge 930 server with four hyper-threaded 18-core Intel Xeon CPUs and a total of 144 parallel threads. The system memory was 1.25 TB, but the critical point for the success of the calculation was the performance of the disks. Swiss electronics distributor Brack.ch supplied twenty-four 6 TB hard drives to hold the vast amounts of data generated during the calculation and backups of intermediate results. Always eager to support their customers in special projects, Brack.ch recommended Seagate Enterprise Capacity hard disks for their high data transfer rates and 24/7 reliability. In the end, more than 7 PB of data were read and written over the course of the project, filling each hard drive sixty times over without a glitch.

The calculation of pi with Alexander J. Yee's highly parallelized program γ-cruncher put a heavier strain on our hardware than even macromolecular crystallography experiments with the largest EIGER detectors. Swapping of pages from memory to disk caused high sustained data transfer rates, while the server ran at full CPU load for extended periods of time. "While the hard drives were specifically chosen for this test, it revealed the kind of system reliability that our customers have come to expect of DECTRIS systems", says Stefan Brandstetter, Head of Product Management at DECTRIS. The calculation proceeded without unexpected interruption and finished in 105 days.

With this calculation, DECTRIS has not only set a new world record for the precision of pi but, says Clemens Schulze-Briese, CSO of DECTRIS, also "performed a veritable stress test on a system similar to those some of our customers will encounter as DCU of their latest EIGER X 9M and EIGER X 16 detectors". The server has passed this test with flying colors, to the benefit of mathematicians and crystallographers alike.

If you are interested in obtaining all 22.4 trillion digits of pi, please contact info@dectris.com. Peter Trüb describes the pi project in more detail in his personal blog and has published his analysis of the digits at ArXiv. Pi appears normal.

Ultrafine phi-slicing maximizes EIGER MX data quality

EIGER data are superior to PILATUS data

Villigen, 6 Sep 2016. The MX group at the Swiss Light Source has just published a paper that describes the best way of collecting macromolecular diffraction data with EIGER. In collaborative work with scientists at DECTRIS, Meitian Wang and colleagues show that measuring with EIGER leads to higher-quality data than with PILATUS, as long as the data are finely enough sliced in phi. Whereas an oscillation angle of 1/2 the XDS mosaicity was recommended for PILATUS, EIGER data benefit from collection at 1/10 the XDS mosaicity. Regardless of phi-slicing, the smaller pixel size of EIGER further increases data quality in the highest resolution shells because less background is measured with weak reflections.

tl_files/root/news/0_Success_Stories_2016/casanas2016_Fig4a_sm.jpg
Smaller pixels of EIGER lead to increased data quality. When EIGER data are binned fourfold, data quality as indicated by Rmeas decreases, in particular for higher-resolution information.

The paper, published in the September 2016 issue of Acta Crystallographica Section D and freely available under a Creative Commons Attribution License, marks another milestone in the progress of macromolecular crystallography. In 2006, PILATUS introduced Hybrid Photon Counting and revolutionized data collection, both in terms of speed and in terms of quality. Now, ten years later, EIGER takes science further yet. Ultrafine phi-slicing provides data of unprecedented quality, collected even more quickly than before. Like those researchers that recently solved structures of Zika virus envelope protein in complex with a neutralizing antibody and of CRISPR-Cpf1 in complex with guide RNA and target DNA from EIGER data, crystallographers will determine structures from ever more challenging crystals and increase our understanding of complex biological processes in health and disease.

Seeing progress towards a vaccine against Zika virus

EIGER X 9M debuts in a project of great medical importance

Baden, 4 Aug 2016. As the Zika epidemic spreads through the American continent and threatens not only this month's Olympic Games in Rio de Janeiro but also, more insidiously, the health of millions, exceptional efforts go towards finding treatments or prevention of Zika infection. Today, the journal Nature published an article that reports a breakthrough on the way towards a vaccine against Zika (Barba-Spaeth et al.). The results critically depended on X-ray crystallographic data collected at Synchrotron SOLEIL with DECTRIS detectors. For beamline Proxima-2A, it is the first publication with data from their new EIGER X 9M detector, but it is also the first structure in the Protein Data Bank solved with EIGER X 9M data.

DECTRIS is excited that EIGER X 9M premieres as part of a research project of such significance and hopes that the findings are quickly translated into medical solutions. We congratulate Bill Shephard and his team at the beamline as well as the group around Felix Rey at Institut Pasteur, Paris, on this accomplishment.

tl_files/root/news/0_Success_Stories_2016/zika_e_AB.png
Figure 1. Neutralizing antibodies bind to Zika virus envelope protein. The dimeric Zika virus envelope protein is shown at the bottom (blue and cyan). Bound to it on the left is the antigen-binding fragment (Fab) of a neutralizing antibody crossreactive against dengue (heavy and light chain, brown and pale orange, respectively). The accession code for the structure is 5lcv.

Zika virus is a flavivirus closely related to the causative agents of dengue and yellow fever and like those, it can be communicated sexually or transmitted by Aedes mosquitoes. The virus is linked to blindness, deafness, seizures and a form of temporary paralysis called Guillain-Barré syndrome, but the majority of infected adults experience no symptoms. Children born to infected mothers, on the other hand, show a high incidence of microcephaly, a birth defect characterized by abnormally small heads and irreversible brain damage. The connection between Zika infection and microcephaly was recently shown experimentally in flies (Cugola et al.) and is now generally accepted to be true in humans as well.

Since its appearance in Brazil a year ago, Zika has radiated outward on the American continent to such an extent that the World Health Organization has declared it a Public Health Emergency of International Concern. The W.H.O. expects the virus to become established in areas ranging from northern Argentina to the southern United States by the end of 2016, infecting millions of people.

There is currently no treatment of Zika infection and no vaccine exists to prevent it. Indeed, the study of Zika, despite the identification of the virus in Africa nearly 70 years ago, is still in its infancy. Only this year and undoubtedly driven by the threat of vast numbers of new cases has structural information of the virus become available. These recent results dramatically advance our understanding of the virus and open possibilities for the treatment and prevention of infection.

To start with, two structures of the intact virus, solved by electron cryo-microscopy (Sirohi et al., Kostyuchenko et al.), gave a first detailed view of the entire virus. Two more papers reported structures of individual protein components of Zika, envelope protein (Dai et al.) and a fragment of non-structural protein 1 (NS1, Song et al.). The first crystal structure revealed the position of a protective antibody and identified a loop in the protein that might serve as an epitope for future therapeutic antibodies, though their development is a long way off. The structure of full-length NS1, recently solved from PILATUS3 6M data at Advanced Photon Source (APS), showed marked differences in epitopes among flaviviruses – with ramifications for vaccine design (Brown et al.).

Crystal structures of two enzymes essential for the Zika virus life cycle lay the foundation for the rational design of antiviral agents.  The structure of NS3 helicase was determined independently at Shanghai Synchrotron Radiation Facility, with data collected on a PILATUS3 6M detector, and at APS  (Tian et al. and Jain et al.). Helicases from other viruses have already been used as targets for the development of antiviral agents. The structures of NS3 helicase pave the way for similar work in Zika.  Of comparable importance is the structure of Zika protease in complex with an inhibitor, solved from data collected with a PILATUS 6M at PETRA III (Lei et al.).

The intense worldwide structural work on Zika reached a provisional conclusion today with a paper published in Nature that reports the structures of Zika virus E protein, the main target of neutralizing antibodies in flaviviruses, in complexes with antibodies cross-reactive against Zika and a close relative, Dengue (Barba-Spaeth et al.). The antibodies were isolated from dengue patients and shown to neutralize Zika virus. Knowledge of the structure of the epitope-antibody interaction raises the prospect of epitope-focused design of a vaccine that is active against both Zika and dengue viruses simultaneously. Of more immediate impact is the suggestion that one of the two antibodies under study could be used for immune prophylaxis in pregnant women at risk of contracting Zika.

This breakthrough is the result of a broad international collaboration. Clinicians, virologists, immunologists and infectious disease specialists from Paris, London and Vienna and as far away as Thailand and Tahiti combined their efforts, but at the heart of the study are the three crystal structures. The data leading to these structures were acquired at Synchrotron SOLEIL with PILATUS 6M and EIGER X 9M detectors.

While PILATUS 6M has been at the heart of X-ray crystallography for nearly ten years, EIGER X 9M makes its debut in the Protein Data Bank (PDB) with the Zika E protein structure. With this, it follows hot on the heels of EIGER X 4M and EIGER X 16M whose first output was released by the PDB in April and May. EIGER detectors are now firmly established as powerful tools for structural biology and important players in the progress of medicine.