B UILDING ON THE DNA REVOLUTION

academic ability, and to home in on that,” considerations are al- says Fuller. “There’s a history of people so gaining promi-

ECTION with no qualifications who are now senior.” nence. For instance,

S 25 years ago MRC Past as prologue didn’t bother to patent At the end of April, hundreds of former LMB Milstein’s technique researchers will converge on Cambridge to for making monoclon-

PECIAL celebrate the 50th anniversary of Watson and al antibodies, now a S Crick’s DNA paper. They include numerous fundamental tool in Nobel laureates whose prizewinning research many industries. The came after their time at LMB, as well as same was true of prominent department heads, institute direc- Sanger’s sequencing tors, and journal editors. There is no doubt in technology. Today, their minds that LMB is unique. “I don’t patenting is encour- think if you had put the same people in a U.S. aged, says Henderson, institution that they would have done as and several compa- well,” says Rubin. Biological incubator. Hundreds of budding molecular biologists got nies, such as Celltech, But can it continue to be so special? their start at the Laboratory of Molecular Biology, opened in 1962. are associated with Thirty years ago, “the field was much the lab. smaller. It was the place for U.S. postdocs to To keep pace with the burgeoning sci- Klug and Henderson suspect that the go, and the best went,” Rubin explains. entists and staff—about 400, more than place is good for at least a couple of more “Now there are many good places.” Al- twice the number 30 years ago—the build- Nobels. Even today, with universities, though funds still flow relatively freely, pa- ing has doubled in size every decade since medical foundations, and other organiza- perwork, regulations, and other constraints 1962. A new building is in the works. Says tions working to create hotbeds of scien- have crept in, Henderson notes. And while Klug, “I am worried that we will get too tific creativity, LMB still earns strong ku- he and his colleagues pride themselves on big and lose the ethos on which the lab has dos. Says Yale’s Joan Steitz: “There have their small labs, which range in size from 1 been built.” been very good research institutions that to 10 people, they worry that they will fall LMB now relies on a glossy annual re- have tried to capture the flavor and spirit, behind. “There’s so much more you can do port rather than word of mouth to publi- but they haven’t got it.”

with more manpower,” says Pelham. cize its accomplishments. Commercial –ELIZABETH PENNISI OF MOLECULAR BIOLOGY;AND SOCIETY MRC LABORATORY SCIENCE CREDITS: BOTTOM) TO (TOP LIBRARY PHOTO

NEWS DNA’s Cast of Thousands

Watson and Crick’s discovery revealed much, suggested more, but left many details unanswered. Ever since, researchers have been discovering the proteins that unlock DNA and the genetic code

When and Francis Crick with proteins. But rebuilt today, Watson and elucidated the structure of DNA, they dis- Crick’s bare-bones model would be draped covered an elegantly simple molecule. with proteins that kink and curl, repair, and With cardboard cutouts, metal, and wire, otherwise animate DNA. they showed how DNA’s two chains wound around each other, with the paired bases DNA ascendant inside, one full rotation every 10 bases. The age of DNA began well before Crick Their model immediately suggested how and Watson were born. In the 1860s, Image not DNA copied itself and enabled genetic Friedrich Miescher, a Swiss working in information to flow from one generation Tübingen, Germany, isolated a strange, available for to the next. They boasted that they had phosphorus-rich material from the cell nu- online use. found the “secret of life”—essentially, bi- cleus. Within decades, it was clear that this ology’s master molecule that controlled peculiar substance—later identified as nu- the fate of the cell and, consequently, of cleic acids—was fundamental to the cell’s the organism. chemistry. Somehow. Fifty years of research since then has Throughout the early part of the 20th shown that, despite its precision design, this century, biochemists argued about DNA’s molecule can’t dance without a team of cho- role. Some postulated that it was the stuff of reographers. Like a puppet, DNA comes genes; others insisted that proteins carried alive only when numerous proteins pull its “strings.” At the time of their discovery, Naked DNA. Watson and Crick’s first model of Watson and Crick had only the haziest of DNA didn’t begin to reveal the complex set of ideas about how this double helix interacted proteins the molecule needs to do its job.

282 11 APRIL 2003 VOL 300 SCIENCE www.sciencemag.org B UILDING ON THE DNA REVOLUTION S PECIAL California Insti- errors, the cell calls in its molecular repairers. tute of Technolo- These enzymes mark the bad DNA, cut it gy in Pasadena out, and replace it with the right code. One of

showed this to be the best-studied examples is that of bacteria S the case. as they recover from exposure to ultraviolet ECTION Shortly after- light. First, a complex of UvrABC proteins ward, Arthur Ko- recognizes the damage. Then the UvrABC rnberg of Stanford enzyme cuts at two sites a few bases to either University and his side of the defective DNA and whisks away colleagues dem- that piece. DNA polymerase then fills in the onstrated that an gap with the correct bases. enzyme they had discovered several DNA’s messenger years earlier or- Watson and Crick’s discovery left wide open chestrates the syn- the question of how DNA specifies which thesis of each new proteins are made. It was more than a DNA strand. The decade before the “code” itself was worked Chameleon. Although the B-DNA is most common and the one first described, enzyme, DNA out, along with all the intricate details of the certain conditions force this molecule into A- or Z-DNA configurations. polymerase, adds gene-to-protein transition. just the right nu- The first clues that genes specified the genetic code. Even though Oswald Av- cleotide base to the separated DNA strands, amino acids came from V. M. Ingram of the ery, Colin MacLeod, and Maclyn McCarty making sure the new one exactly matches its University of Cambridge. He of Rockefeller University in template. More than 20 additional proteins studied the sickle cell trait, demonstrated in 1944 that DNA and not are also known to perform distinct func- in which two defective proteins carried the genetic code, the debate tions in copying DNA. Some help un- continued; even Crick and Watson at first wind DNA, for example; others mark First glimpse. This 84 (1996) disagreed on this point. the starting point of replication. And x-ray diffraction pat-

CELL , Soon after Watson joined him at the Uni- since then researchers have found at tern hinted that DNA versity of Cambridge, U.K., in 1951, Crick least two more DNA polymerases: ET AL. was helical, thereby was persuaded. Across two continents, they one specializes in making new DNA; helping Watson and and others set out to discover just what another helps repair damaged DNA. Crick come up with DNA looked like. Tapping a newly devel- Some mistakes are introduced long the right structure. oped imaging technique called x-ray crystal- after DNA polymerase has finished its lography, Rosalind Franklin and Maurice job—for instance, when radiation or toxic genes can lead to severe ane- Wilkins of King’s College in London pro- chemicals cause the wrong base or bases to mia, while one causes just mild problems. In (1992) ; S. FROM ADAPTED KIM 9

6 duced images that showed DNA was heli- be substituted or others to be deleted al- 1957, he tracked the defect down to a single

CELL cal. Others were busy envisioning together. This necessitates a amino acid change, a mistake in whatever ., how DNA’s bases were more extensive repair sys- specified the order of amino acids in the he- ET AL arranged to enable it to tem. Faced with such moglobin protein. The work suggested that function. Watson and DNA was that template. Crick’s discovery set- Duplicating DNA. The ring section of Then, in 1961, Marshall tled once and for all DNA polymerase helps this enzyme’s Nirenberg of the U.S. Na- that genes were core work more efficiently as it car- tional Institutes of Health in made of DNA. But ries out DNA replication. The crys- Bethesda, Maryland, real- it took eight more tal structure shows the ring (red ized that a three-base se- years—and the and yellow) sliding down the DNA. quence in DNA, UUU, spec- efforts of many ified the amino acid phenyl- researchers—to alanine; soon the rest of the crack the genetic triplets were deciphered. code contained in the We now know that DNA’s nucleotide bases. triplet code is transmitted Watson and Crick, par- through an intermediary ticularly Crick, had many ideas called messenger RNA STRYER, ED. (W. H.AND CO., FREEMAN 1995); JEREMY NORMAN; X-P KONG .

,L about how DNA worked, something (mRNA), a closely related, their landmark 1953 paper hinted at in its single-stranded nucleic acid. last sentence: “It has not escaped our notice The mRNA ferries this in-

BIOCHEMISTRY that the specific pairing we have postulated formation out of the nucleus immediately suggests a possible copying to the ribosome, which mechanism for the genetic material.” The builds the protein one amino idea was that, as the double helix uncoiled, acid at a time. each strand of an existing DNA molecule Early on, there were clues could act as a template for building another that a short-lived molecule copy of the molecule. In the late 1950s, might be involved. The prime

CREDITS: (TOP TO BOTTOM) FROM FROM CREDITS: BOTTOM) TO (TOP and Franklin Stahl of the suspect was RNA, because it

www.sciencemag.org SCIENCE VOL 300 11 APRIL 2003 283 B UILDING ON THE DNA REVOLUTION

appears in large quantities out- side the nucleus where proteins Reading the code. To transfer a gene’s coding data into protein production,

ECTION are being made by the ribosome. an enzyme called RNA polymerase S But researchers had no idea how (peach) latches onto DNA, unwinds the RNA got there or what it re- the double helix, and begins to build ally did. Then, in 1960, four messenger RNA based on a template groups, of strand of the DNA (top). It moves PECIAL Memorial Sloan-Kettering Can- along the DNA as the nascent RNA S cer Center in New York City and forms (bottom). The crystal structure Sam Weiss of the University of shows the RNA polymerase (gray) Chicago among them, independ- with its clamp (orange) building the ently discovered that the enzyme RNA based on the DNA strand (blue). RNA polymerase strings togeth- er the RNA’s bases much as DNA poly- merase does for DNA. A year later, Sol Spiegelman of the University of Illinois, Urbana-Champaign, showed that RNA poly- DNA’s advisers merase reads this code from the DNA tem- The entire DNA code is not expressed all at RCSB plate, providing one of the strongest clues once. The human’s 31,000 or so genes are ;

that RNA was somehow involved with DNA. turned on and off, singly and in combina- ET AL. At about the same time in Paris, the Pas- tion, depending on which suites of proteins ODISH

teur Institute’s François Jacob and Jacques are needed for specific cell functions. Few ,L Monod proposed that a short-lived messen- researchers gave much thought to the idea ger molecule shuttled DNA’s coding infor- that proteins might regulate genetic activity mation from the nucleus to the cytoplasm. until 1961. In the same paper in which Working with Meselson and Sydney Bren- Monod and Jacob suggested the existence ner at the Laboratory of Molecular Biology, of mRNA, they proposed that cells also Jacob then verified the idea. Thus, by the have regulatory elements that affect gene MOLECULAR CELL BIOLOGY early 1960s, there was little doubt that expression. Then, in 1967, mRNA linked the gene to a protein’s pro- and Benno Müller-Hill of Harvard Univer- duction and that RNA polymerase was cen- much more efficiently when it links up with sity isolated a protein that bound to DNA tral to this process. a protein, called cAMP-CAP. And in the and repressed gene activity. Independently, Again, the process is proving to be even more complex eukaryotes, a whole series of Harvard’s isolated another 1876 (2001); FROM ADAPTED , 2

more complicated than researchers initially steps precedes RNA polymerase activity. transcription factor, as these proteins are 9 2 realized. It turns out that there are three Other proteins and protein-RNA complexes now called. Scores more have since been

RNA polymerases: one for protein-coding are needed to process the RNA that peels off discovered. Some latch onto DNA where SCIENCE genes and two for genes that code for the DNA before it’s ready to exit the the synthesis of mRNA begins, to either , that are never translated into proteins. In ad- nucleus. A key change is the removal of suppress or stimulate gene activity; others ET AL. dition, RNA polymerase has help. For in- noncoding bases transcribed from the non- work from afar. stance, in bacteria, RNA polymerase works coding sections of DNA. Over time, researchers have come to re- alize that multiple proteins, in- Simple to complex. Researchers first teracting in different ways, ex- thought individual proteins latched onto ert exquisite control over gene and regulated DNA activity (W-shaped expression. The same factor ribbon). But now they realize that these can alternate between activat- proteins rarely act alone. Varying combina- ing and repressing transcrip- tions of transcription factors, along with tion, depending on its protein (W. H. FREEMAN AND CO., 2000); A. L. GNATT protein activators or repressors, exert dy- partners. For example, in nor- ET AL. namic and finely tuned control over gene mal human colon cells a pro- ODISH expression. tein called groucho links with ,L a transcription factor called Lef/Fcf and keeps an onco- gene called Myc quiet. But if β-catenin takes the place of

groucho, Myc is turned on. MOLECULAR CELL BIOLOGY Transcription factors often make their mark by bending the DNA so that the enzymes that translate the DNA code into protein can position themselves at the right place on DNA and still interact with one another. For exam-

ple, in a prerequisite for ex- CREDITS: ADAPTED FROM TO BOTTOM) (TOP

284 11 APRIL 2003 VOL 300 SCIENCE www.sciencemag.org B UILDING ON THE DNA REVOLUTION S PECIAL pression of most eukaryotic genes, the 1974, Stanford’s Roger Kornberg (son of ity, whereas the removal of a histone’s acetyl transcription factor TFIID causes the DNA Arthur) proposed that chromatin was quite groups silences nearby genes. molecule to bend, paving the way for other structured—made up of repeating units, Today, that cast has expanded to include

transcription factors. each containing 200 base pairs of DNA enzymes that add and remove methyl S Often, transcription factors work in com- wrapped around pairs of two of four differ- groups and others that do the same with ECTION bination. A half-dozen factors link together ent histones. Those units are now called nu- phosphates. Histones are becoming such to activate the β-interferon gene, for in- cleosomes, and they pack and organize stance. Their association is facilitated by DNA so that under an electron microscope, several copies of yet another protein called it looks like beads on a string. HMGI, which causes DNA to bend sharply For the next 20 years, few re- so that the various transcription factors align searchers thought histones were any- themselves elbow to elbow as they work. thing more than structural sup- And four varieties of specific transcription ports. Because the nucleosomes factors together with more than a half-dozen remained intact during transcrip- other proteins are required to switch on the tion, they didn’t seem to be in- TTR genes in liver cells, so they make that volved in gene regulation. But it blood protein. turns out that although the nu-

Twirled around. Thanks to an eight-protein complex, DNA’s double helix is twisted into .; T. J. RICHMOND a “beads on a string” array. Called histones, those proteins make up the nucleosome, ET AL and 200 bases wind around each nucleosome, then loop to the next. Once considered no more than scaffolding, nucleosomes help control gene activity when groups of en- ODISH ,L zymes such as deacetylation complexes chemically modify the histones (illustration). The crystal structure (top right) reveals DNA (white) encircling various histones, and the bird’s-eye view (bottom left) shows those histones with DNA in peach and green.

cleosome remains intact, the prominent players in DNA activity that two histones need to loosen their grip researchers—Thomas Jenuwein of the Re- MOLECULAR CELL BIOLOGY on DNA for transcription factors to search Institute of Molecular Pathology in gain access. Otherwise, RNA poly- Vienna, Austria, and C. David Allis of the merase has a hard time getting into po- University of Virginia Health Science Cen- sition, and transcription is hampered. ter in Charlottesville—now argue that there

(1997); FROM ADAPTED As early as 1964, Vincent Allfrey of is a “histone code” as complex and impor-

9 8

3 DNA’s attire Rockefeller University realized that histones tant as the DNA code, one that fine-tunes At the time of Watson and Crick’s discov- were often chemically modified by the addi- gene activity and adds more depth to the in-

NATURE , ery, it was already clear that DNA was not tion of many acetyl side groups, which formation encoded in the genes. The idea is

ET AL. really naked in the cell nucleus but was seemed to cause them to slacken their hold slowly catching on, leaving pioneering mo- adorned with proteins. For decades, these on DNA. The observation was all but for- lecular biologists to shake their heads. proteins were considered mere dressing. gotten, however, until 1996, when a slew of Forty years ago, Brenner and others were Indeed, 50 years later researchers are still researchers discovered that histones, too, are convinced that the central questions in mo- figuring out how they interact to regulate also puppets: Various proteins cause them to lecular biology would be answered well be- gene expression. change shape, which in turn alters gene ac- fore the turn of the century. Now they know In the cell nucleus, those proteins— tivity. In a matter of months, the cast of pup- better. The nature of the histone code is just primarily histones—together with DNA peteers included four enzymes that add one of many problems whose complexities make up a complex called chromatin, so acetyl groups to histones and five that re- are left to be unraveled.

CREDITS: (TOP TO BOTTOM) FROM K. FROM CREDITS: BOTTOM) TO LUGER (TOP named because of how it stains in cells. In move them. Acetylation prompts gene activ- –ELIZABETH PENNISI

www.sciencemag.org SCIENCE VOL 300 11 APRIL 2003 285