618 nature structural biology • synchrotron supplement • august 1998

synchrotron supplement

Two techniques have been the pillars ofour burgeoning understanding of biologyat the molecular level: recombinant DNAtechnology and X-ray crystallography.Much as crystal structure determinationbecame a handmaid of chemistry so nowwe see structure determination as impor-tant for underpinning molecular biology.Furthermore, during the last decade — aresult of its excellent collimation, bril-liance, and polychromaticity — synchro-tron radiation has become anindispensable X-ray source for proteincrystallography and therefore an integralpart of modern biological research.

However, muscle contraction, not pro-tein crystallography motivated the use ofsynchrotron radiation as a source for X-raydiffraction. Skeletal muscle contracts by themutual sliding of thick myosin and thinactin filaments. H. E. Huxley first noted thata detailed low angle X-ray fiber diffraction

pattern could be obtained from living frogmuscle. Later it was shown that the lowangle pattern was responsive to the confor-mation of the myosin ‘cross bridges’1 whichprotrude from the thick filaments and movethe thin filaments (the ‘swinging cross-bridge hypothesis’2). Efforts to observe thecross bridges in different conformationsduring a contraction3 showed that muchstronger X-ray sources than could beobtained from rotation anode X-ray tubeswould be required.

In 1963, a group measured the soft X-rays that were emitted from the 1 Gevsynchrotron at Frascati4. At this time anumber of more energetic electron syn-chrotrons were being commissioned, andSchwinger’s theoretical studies5 allowedone to calculate that a 6 Gev machinewould be a powerful X-ray source. In 1964I contacted the DESY (DeutschesElektronen-Synchrotron) with a view to

using DESY as an X-ray source. However,it was only after moving to the Max PlanckInstitute in Heidelberg, Germany in 1968where I was joined by Gerd Rosenbaumthat this idea became a reality. Rosenbaumcontributed essential hands-on experi-ence, acquired at the vacuum-UV syn-chrotron radiation facility (F41) at DESY.Together with Jean Witz we set up acurved quartz monochromator on theDESY vacuum-UV beam-line and inAugust, 1970 we used the resulting mono-chromatic beam to obtain a smudgy dif-fraction picture of the 220 Å equatorialBragg reflection from a sample of insectflight muscle6 (Fig. 1a).

On the basis of this successful trial wewere strongly encouraged by the directorsof DESY to build a laboratory on the syn-chrotron DESY (Bunker-2) for X-ray dif-fraction. This took place in 1971. It wasclear that the constant-current storage

From rare to routineKenneth C. Holmes

The growing importance of synchrotron radiation for structural biology can be charted from the construction anduse of an X-ray beam line at DESY in Hamburg, Germany in 1970, to the completion of the three third generationsynchrotrons in France, the USA and Japan in the 1990s.

Fig. 1 a, The first X-ray diffraction pattern taken with synchrotron radiation; thesample was of insect flight muscle and the picture shows the 220 Å equatorial Braggreflection. b, The first X-ray beam line; it was built in Bunker-2 on the DESY synchro-tron7. The nozzle carrying the berrilium exit window from the vacuum pipe connect-ing with the tangent point in a bending magnet (about 40 m distant) can be seen inthe bottom right of the picture. The plastic box was helium-filled and contained twobent quartz mirrors and defining slits and associated motors. The Guinier bentquartz crystal monochromator was housed in an evacuated pot (center of picture). Avacuum tube led to the specimen holder (in the author’s right hand) and the guardslits. A film or linear detector could be mounted in the focal plane (rear). Because ofhigh radiation levels in the bunker all adjustments had to be controlled remotely.The beam-line was equipped with TV camera for remote monitoring.

ba

nature structural biology • synchrotron supplement • august 1998 619

synchrotron supplement

ring DORIS (at this time still being built)would be a much better source than thesynchrotron, which dumped its beam 50times a second, but in the mean timemuch useful experience was gained usingDESY. The first X-ray beam line (Fig. 1b)was operational by the summer of 19727,allowing a number of experiments oninsect muscle and other biological materi-als to be performed8. The long awaitedtime-resolved experiments on frog muscleneeded the strength of the storage ringand were finally carried out by H.E.Huxley and his team on DORIS inBunker-4 in 19809.

While the high energy physics commu-nity was very supportive one could notoverlook the fact that they were payingfor the machines and we were usingthem, almost parasitically. They hadtheir own priorities concerning energyand modes of operation, which often ledto long periods with no useful beam.Furthermore, the machines were neitherparticularly stable nor reliable. Initialattempts by our laboratory to use the X-ray beam line in Bunker-2 for X-raycrystallography led to a distinct improve-ment over a rotating anode tube (~10-fold gain) but when the workingconditions were considered it almost didnot seem worth the bother10. However,experiments carried out about the sametime at Stanford on the constant-currentstorage ring SPEAR showed a ~50-foldgain and demonstrated the power of themethod11. Moreover, the fantastic colli-mation of the synchrotron beam beganto make its impact: accurate data couldbe obtained from large unit cells andsmall crystals.

From its inception the EuropeanMolecular Biology Laboratory under itsdirector general John Kendrew was verysupportive of the synchrotron radiationenterprise and in 1975 Bunker-2 and asecond laboratory on the storage ringDORIS (Bunker-4) became the EMBLoutstation at DESY. X-ray crystallogra-phy in the outstation grew slowly butsteadily. Low angle scattering and fiberdiffraction were still supported but losttheir pre-eminence. Furthermore, in thecourse of time the high energy physicistsmoved on to bigger machines, leavingthe synchrotron radiation user commu-nity in sole charge. Synchrotron radia-tion had become important enough tocontemplate building dedicatedmachines that would be run for the con-

venience of synchrotron radiation usersrather than high energy physics. The useof synchrotron radiation as an X-raysource for macromolecular crystallogra-phy took off and even the small moleculecrystallographers began to take an interest.

In the last decade a number of technicaladvances have helped make synchrotronradiation widely available as a standardtechnique for macromolecular crystallog-raphy (see articles by Keith Wilson andRobert Sweet in this supplement). Clearly,the great improvements in stability andbrilliance of synchrotron radiationsources have played an important role.Since our first experiments at DESY thebrilliance has risen by more than fiveorders of magnitude. Additionally, cryo-crystallography has made it much easier touse synchrotron radiation for experi-ments. In 1934 J.D. Bernal founded pro-tein crystallography by photographingpepsin crystals wet sealed in a capillarytube. However, when X-rays hit the moth-er liquor it becomes a soup of free radicalsand radiation damage takes out even thehardiest protein crystal in time. A numberof groups experimented with non-aque-ous environments and liquid nitrogentemperatures. It was established thatfrozen crystals could survive if the freezingwas very rapid12. In 1990 Teng13 combinedthis notion with a traditional mounting:catch the crystal in a film in a thin wireloop and plunge it into liquid nitrogen, amethod which is particularly good forsmall crystals. and yields crystals whichare nearly immortal. The use of areadetectors, particularly imaging plates,instead of photographic film to collectaccurate data marks another advance inthe field. Finally Hendrickson’s multi-wavelength anomalous diffraction (MAD)technique14 (see the article by CraigOgata), particularly when used withselenomethionine, is changing the waycrystallographic experiments are per-formed — by providing an almost univer-sal method of experimentally determiningphases (that can be carried out only on asynchrotron source).

Initially, synchrotron X-ray sourceswere considered suitable only for ‘difficult’problems such as determining the struc-tures of viruses. Furthermore, asdescribed above they could be very frus-trating to use. However, the accuracy ofthe data coupled with the unique ability tovary wavelength has led to the routine useof these sources for protein structure

determinations. Moreover, protein struc-tures are no longer solved solely by spe-cialists but rather are increasinglydetermined by cell biologists and bio-chemists. Thus protein crystallographyusing synchrotron radiation is now expe-riencing rapid growth: the annual publi-cation of new structures has grown to aflood of 800 per year (Brookhaven ProteinData Bank 1996) of which most dependon synchrotron radiation. Given theimpact of the genome projects, the magni-tude of future growth and the character offuture research at synchrotron facilitiesaround the world will for some decades belimited by the funding of biological sci-ence and availability of beam lines ratherthan by scientific need, which is virtuallyopen-ended.

Kenneth C. Holmes is at the Max PlanckInstitut für medizinische Forschung69120 Heidelberg, Germany. email:holmes@otto.mpimf-heidelberg.mpg.de

1. Reedy, M.K., Holmes, K.C. & Tregear, R.T. Inducedchanges in the orientation of the cross-bridges ofglycerinated insect flight muscle. Nature 207,1276–1280 (1965).

2. Huxley, H.E. The mechanism of muscularcontraction. Science 164, 1356–1366 (1969).

3. Huxley, H.E. & Brown, W. The low angle X-raydiagram of vertebrate striated muscle and itsbehaviour during contraction and rigor. J. Mol.Biol. 30, 383–434 (1967).

4. Cauchois, Y., Bonnelle, C. & Missoni, G.Rayonnement électromagnétique — Premiersspectres X du rayonnement d’orbite du synchrotronde Frascati. Compte-rendu hebd. Séanc. Acad. Sci.Paris 257, 409–412 (1963).

5. Schwinger, J. On the classical radiation of acceleratedelectrons. Phys. Rev. 12, 1912–1925 (1949).

6. Rosenbaum, G., Holmes, K.C. & Witz, J. Synchrotronradiation as a source for X-ray diffraction. Nature230, 434–437 (1971).

7. Barrington Leigh, J. & Rosenbaum, G. A report onthe application of synchroton radiaton to low-angle scattering. J. Appl. Crystallogr. 7, 117–121(1974).

8. Barrington Leigh, J. & Rosenbaum, G. SynchrotronX-ray sources: a new tool in biological structuraland kinetic analysis. Annu. Rev. Biophys. Bioeng. 5,239–270 (1976).

9. Huxley, H.E. et al. Millisecond time-resolvedchanges in X-ray reflections from contractingmuscle during rapid mechanical transients,recorded using synchrotron radiation. Proc. Natl.Acad. Sci. USA 78, 2297–2301 (1981).

10. Harmsen, A., Leberman, R. & Schulz, G.E.Comparison of protein crystal diffraction patternsand absolute intensities from synchrotron andconventional X-ray sources. J. Mol. Biol. 104,311–314 (1976).

11. Phillips, J.C., Wlodawer, A., Yevitz, M.M. &Hodgson, K.O. Applications of synchrotronradiation to protein crystallography: preliminaryresults. Proc. Natl. Acad. Sci. USA 73, 128–132(1976).

12. Garman, E., F. & Scheider, T.R. Macromolecularcrystallography. J. Appl. Crystallogr. 30, 211–237(1997).

13. Teng, T.Y. Mounting of crystals for macromolecularcrystallography in a free-standing thin film. J. Appl.Crystallogr. 23, 387–391 (1990).

14. Hendrickson, W. Determination of macromolecularstructures from anomalous diffraction ofsynchrotron radiation. Science 254, 51–58 (1991).