Skip to main content
Premium Trial:

Request an Annual Quote

Synchrotrons Gone Mad


Structural biologists are gaining clout in the land of particle accelerators

By Aaron J. Sender


In his Columbia University office, Wayne Hendrickson carefully sketches bars, loops, and twists for a high-throughput crystallography slide presentation he’ll give the following day in Los Angeles. He is perfectly self-composed, soft-spoken, and elegant. Yet in the world of protein crystallography — the shooting of x-rays at proteins to determine their shape — he is the MAD scientist.

He earned that reputation not for outrageous and illogical flights of fantasy, but for his remarkable contribution to the field of protein analysis: a technique called Multi-wavelength Anomalous Diffraction that has made high-throughput structural genomics a reality. MAD has fundamentally altered the landscape of protein crystallography and lured biologists by the thousands to synchrotrons — half-billion-dollar, kilometer-circumference structures that act as x-ray machines on steroids.

When he first developed MAD in the mid-1980s, “most people viewed it as pretty esoteric,” says Hendrickson. The method required the brilliant radiation and the tuning of x-ray wavelengths only possible at a synchrotron — a radiation source then foreign to most biologists. “It was known, but not used very much,” says Hendrickson. “It was not very accessible.”

Synchrotrons had long been the domain of physicists and material scientists. The few biologists — pioneers such as Hendrickson and Keith Hodgson (now director of the Stanford synchrotron) — brave enough to venture into their territory were often viewed as parasites.

But in the last decade a handful of new, higher-energy synchrotrons have come on-line and biologists are now a dominant force. At the four DOE-funded synchrotrons, for example, biologists made up only six percent of users in 1990. Today nearly half the users are biologists. The numbers are similar worldwide.

Beaming Biologists

The effect of the introduction of MAD beamlines (facilities built around the doughnut-shaped synchrotron optimized for MAD experiments) on protein structure determination has been profound. In 2001 more than 70 percent of structures published will use synchrotron radiation compared to only 18 percent in 1993, before the new beamlines began to operate. During the same time, the number of deposits in the Protein Data Bank has jumped from fewer than 2,000 to nearly 16,000 structures.

Before MAD, crystallographers had to crystallize different versions of a protein. The MAD method uses one crystal and instead compares the diffraction patterns at various wavelengths, eliminating inaccuracies introduced from data of different crystals and shortening the often months-long effort of growing extra crystals. Hendrickson also created a method to systematically introduce a tag, the element selenium, into proteins. The previous technique of getting heavy metals into the protein was a matter of trial and error.

“What MAD experiments have meant is that you can solve structures in an hour, whereas previously it would take many visits over days and weeks,” says Sean McSweeney, who heads the macrocrystallography group at the European Synchrotron Radiation Facility in Grenoble, France. “It’s the most important mechanism for solving new structures.”

Several other developments over the past decade have helped move things along: the construction of the higher energy, third-generation synchrotrons; faster digital detectors; better hardware and software for managing and analyzing the gigabytes of x-ray diffraction data; and even something as simple as methods for freezing crystals so they are not damaged by the powerful radiation. “The whole thing has gotten faster, more convenient, slicker, and more reliable,” says Janet Smith, a protein crystallographer at Purdue University who chairs BioSynch, a synchrotron users group. “It’s no longer such a wacky adventure.”

Getting in Synch

As biologists flock to synchrotrons, the facilities are scrambling to meet the demand. “With the current number of beamlines we only give out about one-third the amount of time requested,” says Thomas Earnest, head of the Berkeley Center for Structural Biology at the Advanced Light Source synchrotron. ALS has recently gone from one protein crystallography beamline to three and plans to have nine running within the next few years. Dozens of other MAD-capable beamlines are in the works around the globe.

Yet demand is still likely to outpace availability. The National Institute of General Medical Science has recently launched a Protein Structure Initiative with $150 million committed to seven consortiums that will soon be clocking countless hours at synchrotrons. A European Union structural genomics effort is set to begin hogging beamtime as well. Commercial high-throughput protein structure players, such as Structural GenomiX, co-founded by Hendrickson, and Syrrx, as well as pharmaceutical companies, are funding their own beamlines to secure access.

Unlike genome sequencing, where a single company such as Celera was able to compete against the public sector, structural genomics forces a symbiotic relation. “The private enterprise has to work with federal government to get the data,” says Gopal Shenoy, senior scientific director for the Argonne National Lab synchrotron. “There is no way you can expect a Structural GenomiX to build its own synchrotron.”

Although it’s not uncommon to hear colleagues associate his name with a certain prestigious prize endowed by a Swedish foundation, Hendrickson shies away from taking complete credit for his influence on protein crystallography. “At some level I suppose it’s flattering to have people associate the methodology with yourself,” he says. “To some extent it’s overblown.”

Hendrickson, who says his innovation was “incremental” and “two percent inspiration, 99 percent perspiration,” is now witnessing his vision come to fruition. He relates a conversation in the mid-’80s with Nobel laureate Hamilton Smith, now at Celera, about the prospects of protein crystallography becoming a routine biological tool. “He was asking how easily that would be done. My response was that I could imagine in the near future we would be able to provide procedures so that anybody could do it,” says Hendrickson. “The near future is here now.”

On the following pages, GT takes a closer look at several synchrotrons gone MAD.

Biological Beamline Boom

“When we conceived the synchrotron back in the ’80s we never thought that protein crystallography was going to be a dominant part of this facility,” says Gopal Shenoy, senior scientific director of the Advanced Photon Source that’s been operating since 1996 at Argonne National Laboratory outside of Chicago. “You had to drag protein crystallographers to synchrotrons.”

Boy, have things changed. Forty-five percent of APS’s 3,500 registered users are now biologists. And most of the new construction of experimental stations, each with at least two beamlines, is for protein structure determination. Six are currently up and running and another four are in the works.

The San Diego-based firm Stuctural GenomiX, co-founded by MAD scientist Wayne Hendrickson, for example, has just begun work on its own station here that will include three beamlines.

“What’s interesting is that this corporation’s entire product will come out of APS,” says Shenoy. As on other privately funded beamlines, Structural GenomiX will be required to return 25 percent of the beamtime to general users.

Purdue University protein crystallographer Janet Smith, who regularly logs time at Argonne’s synchrotron, says that with developments such as MAD, resistance to biologists is gone. “Now when we apply for beamtime, no one asks anymore, ‘Do you really need synchrotron radiation to do this experiment?’ It’s like saying, ‘Do you really need color on your computer screen?’ It’s not a question.”

Shenoy, a material scientist by training, points to the biological beamline boom as proof of the value of funding physical science. “Without physical sciences development, biological sciences can’t benefit.” — AS

Advanced Photon Source at Argonne National Laboratory


• At 7 GeV, the highest energy source in the US and second worldwide only to Spring8 in Japan

• 34 sectors, each with two or more beamlines

• 21 operational sectors; six dedicated to protein crystallography; four more to be online within next few years


Sectors in use for structural biology include:

IMCA: funded by a consortium of 12 pharma and chemical companies

BioCars: run by University of Chicago, focuses on structural changes over time

SBC: DOE funded for use by independent investigators

COM: provides access to smaller companies that can’t afford their own sectors. Current users include MediChem and Structural GenomiX.

DND: funded by DuPont, Northwestern University, and Dow


Sectors planned for structural biology:

SER: 20 member institutions, administered by University of Georgia

SGX: Structural GenomiX

NE: funded for the use of a consortium of northeastern institutions, including Columbia, Cornell, Harvard, Memorial Sloan Kettering, Rockefeller, and Yale

GM/CI: in early planning stage for use by NIGMS and NCI grantees

LS: funded by Michigan’s tobacco settlement for use by a consortium primarily of the state’s universities


Global Grenoble

“I’m sitting here looking out the window at two mountain ranges, a valley, and a blue sky,” says Sean McSweeney, head of the macromolecular crystallography group at the European Synchrotron Radiation Facility, sponsored by 15 European countries and Israel and located in Grenoble, France. “We have a lot of people who come here in part for the mountains and the scenery,” he says. “But obviously they come for the scientific life as well.”

In fact, the ESRF, operating at six giga-electron-volts since 1994, produces mountains of biological data. It was the first high-energy, third-generation synchrotron built and has had a beamline dedicated to protein crystallography from day one. Today five beamlines solely attack protein structures. Since the first MAD-capable beamline went online in 1995, ESRF’s x-rays have solved more than 400 published structures.

As technology advances, biologist can use beamtime more efficiently. “About six years ago I did a MAD experiment which lasted eight days,” McSweeney recalls. “There were three of us who worked 24 hours a day in shifts just to keep the experiment going. And then we spent three months analyzing the data,” he says. “Now the beamlines are set up so that a lot is automatic. You would do it in less than half a day.” Still, to keep up, “there’s a need for at least two more protein crystallography beamlines,” says McSweeney.

He points to the way the facility is organized as a key to its productivity. “This will get me in trouble with my American friends,” he says, but unlike US facilities, “because we both own the beamlines and run the synchrotron, we operate the two together and can put the science ahead of the bureaucracy.”

— AS

European Synchrotron Radiation Facility, Grenoble, France

• Online since 1994 at 6 GeV; the oldest high-energy, third-generation synchrotron

• 40 operating beamlines; five dedicated to protein crystallography, with eight more using a portion of their time for solving protein structures

• Gearing up for European Union-funded structural genomics project; plans for creating a structural biology center for high-throughput protein expression, purification, and crystallization in collaboration with EMBL


Synchrotron by the Bay

At 1.9 GeV, the Lawrence Berkeley National Laboratory’s Advanced Light Source, situated in the East Bay Hills overlooking San Francisco Bay, operates at the lowest energy of the world’s four third-generation synchrotrons. Still, it produces light in the x-ray region of the electromagnetic spectrum that is one billion times brighter than the sun.

Although not everybody would agree, Thomas Earnest, head of the Berkeley Center for Structural Biology, which runs the three ALS protein crystallography beamlines, says the facility’s lower energy offers a distinct advantage. “Protein crystals, unlike other materials, are radiation sensitive. There’s a certain dose-rate limit that you can give crystals, ” he says. Besides, he adds, other parameters, such as a high current and stability of the x-ray beam, more than make up for the lower energy. “Especially for MAD experiments, you want good energy resolution. If your beam is moving around, it hits your crystal at different positions at different times,” Earnest says.

Despite the lower energy radiation, biologists still flock to ALS. “When we began on our first beamline in 1997, within a year we had essentially a third of the total number of the users at the ALS — and that’s one beamline versus about 20 at the time,” says Earnest. The synchrotron began operating in 1993, but the first protein crystallography beamline wasn’t built until four years later.

Current plans to increase the number of protein structure beamlines to nine may not be enough, says Earnest. “We really need to increase the efficiency through automation,” he says. “Soon software will actually run the beamlines instead of people.”


Major structural genomics projects include:

Berkeley Structure Genomics Center: one of the seven NIGMS-funded consortia as part of its $150 million Protein Structure Initiative. Members include LBNL, UC Berkeley, Stanford, and UNC-Chapel Hill.

San Diego-based Genomics Institute of the Novartis Research Foundation and its structural genomics company spinoff Syrrx jointly fund their own beamline.

Advanced Light Source at Lawrence Berkeley National Laboratory

• At 1.9 GeV, the lowest energy source of the world’s four third-generation synchrotrons

• 27 operating beamlines; three dedicated to structural biology; five more have already received funding, with one more under serious consideration

• Strong emphasis on automation, from software to robotics


The Scan

Shape of Them All

According to BBC News, researchers have developed a protein structure database that includes much of the human proteome.

For Flu and More

The Wall Street Journal reports that several vaccine developers are working on mRNA-based vaccines for influenza.

To Boost Women

China's Ministry of Science and Technology aims to boost the number of female researchers through a new policy, reports the South China Morning Post.

Science Papers Describe Approach to Predict Chemotherapeutic Response, Role of Transcriptional Noise

In Science this week: neural network to predict chemotherapeutic response in cancer patients, and more.