A series of articles for general audiences Unraveling the Mystery of Protein Folding by W. A. (Bill) Thomasson

This series of essays was that allow our muscles to con- developed as part of FASEB’s tract, and the collagen that makes efforts to educate the general up our tendons and ligaments public, and the legislators whom (and even much of our bones)—all it elects, about the benefits of are proteins. fundamental biomedical To make proteins, ‘machines’ research—particularly how known as ribosomes string to- investment in such research gether amino acids into long, leads to scientific progress, linear chains. Like shoelaces, improved health, and economic these chains loop about each other well-being. in a variety of ways (i.e., they fold). But, as with a shoelace, Alzheimer’s disease. Cystic fibro- only one of these many ways sis. Mad Cow disease. An inher- allows the protein to function ited form of emphysema. Even properly. Yet lack of function is many cancers. Recent discoveries not always the worst scenario. show that all these apparently For just as a hopelessly knotted unrelated diseases result from shoelace could be worse than one protein folding gone wrong. As that won’t stay tied, too much of a though that weren’t enough, many misfolded protein could be worse of the unexpected difficulties than too little of a normally folded biotechnology companies encoun- one. This is because a misfolded ter when trying to produce human protein can actually poison the proteins in bacteria also result cells around it. from something amiss when Early Studies proteins fold. What exactly is this phenom- The importance of protein folding enon? We all learned that pro- has been recognized for many teins are fundamental compo- years. Almost a half-century ago, nents of all living cells: our own, Linus Pauling discovered two quite the bacteria that infect us, the plants and animals we eat. The W. A. (Bill) Thomasson, Ph.D., is a science and medical writer based in Oak Park, IL. hemoglobin that carries oxygen to Jonathan A. King served as science writer. our tissues, the insulin that sig- This series is available on FASEB’s Public nals our bodies to store excess Policy Home Page at sugar, the antibodies that fight http://www.faseb.org/opa/ or as reprints from FASEB’s Office of Public Affairs, 9650 infection, the actin and myosin Rockville Pike, Bethesda, MD 20814. simple, regular arrangements of egg cools, the proteins don’t re- amino acids—the α-helix and the turn to their original shapes. β-sheet (see the box, Fundamental Instead, they form a solid, in- Patterns of Protein Structure) – soluble (but tasty) mass. This is that are found in almost every misfolding. Similarly, biochemists protein. And in the early 1960s, have always cursed the tendency Christian Anfinsen showed that of some proteins to form the in- the proteins actually tie them- soluble lumps in the bottom of selves: If proteins become un- their test tubes. We now know folded, they fold back into proper that these, too, were proteins shape of their own accord; no folded into the wrong shapes. shaper or folder is needed. Until recently, biochemists Of course, neither Pauling nor lacked the tools to study these Anfinsen nor the committees that insoluble lumps. Nor did they awarded them their respective expect such masses would be Nobel prizes knew at the time particularly interesting. The that these discoveries would be so prevailing view at the time was important for understanding that the lumps were just hope- Alzheimer’s disease or cystic lessly tangled and completely fibrosis. And when Pauling, at amorphous masses of protein least, was doing his breakthrough fibers (aggregation). Researchers studies, he could hardly have eventually discovered that these imagined the enormity of today’s aggregates of incorrect folding biotechnology industry. What could be highly structured, but scientists did know is that any before this crucial insight and process that was so fundamental before proper investigative tools to life as protein folding would were developed, biochemists have to be of the utmost practical simply threw their fouled test importance. tubes away. But research did not stop with ‘Gunking Up’ Tissues Pauling and Anfinsen. Indeed, we now know that Anfinsen’s conclu- As far back as the start of this sions needed expansion: Some- century, physicians have been times a protein will fold into a noticing that certain diseases are wrong shape. And some proteins, characterized by extensive pro- aptly named chaperones, keep tein deposits in certain tissues. their target proteins from getting Most of these diseases are rare, off the right folding path (see the but Alzheimer’s is not. It was box, Molecular Chaperones). Alois Alzheimer himself who These two small but important noted the presence of “neurofibril- additions to Anfinsen’s theory lary tangles and neuritic plaque” hold the keys to protein folding in certain regions of his patient’s diseases. brain. Tangles are more or less We’ve known since antiquity common in diseases that feature (but didn’t know we knew) that extensive nerve cell death; plaque, protein folding can go wrong. however, is specific to Alzheimer’s. When we boil an egg, the proteins The major question, which has in the white unfold. But when the only recently been answered, is Fundamental Patterns of Protein Structure More than a half century ago, evidence began to accumulate that a major part of most proteins’ folded structure consists of two regular, highly periodic arrangements, designated α and β. In 1951 researchers worked out the precise nature of these arrangements. The key to both structures is the hydrogen bond. A hydrogen atom is nothing more than a proton with a surrounding electron cloud. When one of these atoms is chemically bonded to an electron-withdrawing atom such as nitrogen or oxygen, much of the electron cloud moves toward the nitrogen or oxygen. The proton is thus left almost bare, with its positive charge largely unshielded. If it comes close to another atom with a bit of ex- tra negative charge (typi- cally, an oxygen or nitrogen atom), the par- tial positive and negative charges will attract each other. It is this attraction that produces the hydrogen bond and stabilizes the α and β structures. The a struc- ture, nw called α-helix, is a right-hand spiral stabilized by hydrogen bonds (A) A model alpha-helix shows the hydrogen bonds (dotted lines) between oxygen and hydrogen atoms of the between each fourth amino acid up the chain. (B) beta-sheets are also held together by hydrogen bonds. The transparent amino acid’s ni- arrows show the direction of individual beta-strands. Chains running in the same direction (left pair) are trogen atom called parallel beta sheet; strands running in opposite directions (right pair) are said to be anti-parallel and the oxygen beta-sheet. The atom coloring is as follows: carbon = green, oxygen = red, nitrogen = blue, and white = atom of the hydrogen. (Courtesy: Stanley Krystek, Bristol-Myers Squibb, Pharmaceutical Research Institute) fourth one up the chain. This means that there are 3.6 amino acids for each turn of the helix. The main part of the amino acid (the side chain, designated “R” in the figure) sticks out from this spiral backbone like the bristles on a bottle brush. The β structure is now called β-sheet. It is essentially flat, with the side chains sticking out on alternate sides. β-sheet is also stabilized by hydrogen bonds between nitrogen and oxygen atoms. In this case, however, the hydrogen-bonded atoms belong to different amino acid chains running alongside each other. The sheets are ‘parallel’ if all the chains run in the same direction and are ‘antiparallel’ if alternate chains run in opposite directions. Antiparallel sheets are often, but not always, formed by a single chain looping back upon itself. When a single chain loops back on itself to form an antiparallel β-sheet, the one to three amino acids linking the two strands are known as a β-turn. Today, scientists recognize the β-turn as one of the funda- mental elements of protein structure. All other local arrangements of amino acids are described as ‘ran- dom coil,’ although they are random only in the sense of not being periodic. whether plaque causes millions of potential folded states Alzheimer’s or, like tangles, is a to choose from, consistently found consequence of it. the correct one—and did so within Further investigation showed seconds to minutes. that neuritic plaque (unrelated to Could there be specific, critical the plaque that clogs atheroscle- intermediates (partially folded rotic blood vessels and causes chains) in the folding process? heart attacks) is composed almost This turned out to be a difficult entirely of a single protein. De- question to answer. Partially posits of large amounts of a folded chains don’t stay that way single, insoluble protein around very long; they become fully the degenerating nerve cells of folded chains in a fraction of a Alzheimer’s disease eventually second. Nevertheless, by the provided a key to understanding early 1980s researchers had not the disorder. only found clear evidence for the It was development of the bio- existence of partially folded pro- technology industry that unex- teins, but also realized the key pectedly spurred interest in in- role these played in the folding soluble protein gunk. This indus- process. try can produce proteins (often One study involved the diffi- otherwise difficult-to-obtain hu- culty in getting bovine growth man proteins) quickly and eco- hormone to fold properly. Al- nomically in bacteria. To their though the unfolded proteins were surprise, however, scientists who not sticky, and the fully folded worked for biotech companies proteins were not sticky, the often found two things: protein partially folded molecules stuck to that was supposed to be soluble each other—a first clue as to the instead precipitated as insoluble origins of misfolded lumps (at inclusion bodies within the bacte- least for purified proteins in test ria and proteins that were sup- tubes). It still remained unclear posed to be secreted into the why misfolding occurred in cells surrounding medium instead got under certain circumstances but stuck at the bacterial cell wall. not under others. This puzzling activity led scien- Temperature Sensitivity tists, almost for the first time, to seriously study just what goes The early 1980s also saw one of wrong during protein folding. the first serious investigations of Further Studies protein misfolding. These stud- ies focused on temperature- In the decades after Anfinsen’s sensitive mutations (mutations work, the National Institutes of allowing growth at 75_F but not Health and the National Science at 100_F) in the tailspike protein Foundation continued to finance of bacteriophage P22. Neither research in several laboratories. bacteriophage P22, a virus that Working in relative obscurity, infects certain bacteria, nor its these protein biochemists tried to tailspike protein has any practi- discover how a completely un- cal importance in themselves. folded protein, with hundreds of Faced with thorny problems, Molecular Chaperones

The beauty and the frustration of science are that they are constantly producing surprises. Almost three decades after Christian Anfinsen had won the Nobel Prize for demonstrating that protein folding is governed solely by the protein itself, other scientists discovered that some proteins have helped in the process. This help consists of proteins called chaperones (or chaperonins) that are associated with the target protein during part of its folding process. However, once folding is complete (or even before) the chaperone will leave its current protein molecule and go on to support the folding of another. Proper folding of some proteins appears to call for not just one chaperone, but several. Especially clear evidence for such multi-step chaperoning is provided by test-tube experiments on a protein known as rhodanese. Proper folding of this pro- tein, the experiments show, requires five different chaperone-type proteins acting at two distinct steps in the operation. Early in the folding process, rhodanese binds to a chaperone known as DnaK; the complex that binds a further chaperone: DnaJ. Somewhat later, a protein known as GrpE catalyzes transfer of the partially folded rhodanese to another chaperone, GroEL, and its partner, GroES. These latter two proteins then see rhodanese all the way through to its properly folded state. Several lines of evidence suggest that chaperones’ primary function may be to prevent aggregation. For example, a chap- erone found in the ‘power plant’ organelles of mammalian cells (but otherwise simi- lar to GroEL) has been shown to consist of 14 protein chains arranged as two dough- nuts stacked on top of each other (see fig- ure). The chaperoned protein sits inside the two doughnut holes, safely sequestered from other molecules with which it might aggregate. A role for chaperones in preventing ag- The molecular surface of the immunodominant heat-shock chaperonin- gregation is also suggested by what hap- 10 of mycobacterium leprae shows the multiple subunits of the protein pens to mammalian proteins produced in complex forming one doughnut layer (Protein data bank entry 1lep). bacteria. Although bacteria have chaper- The second complex would stack together with this complex having the ones, they are not the same as those in chaperoned protein inside the open hole. Each subunit is colored differ- mammals. It is thus easy to imagine that ently. (Courtesy: Stanley Krystek, Bristol-Myers Squibb, Pharmaceuti- they may be relatively ineffective toward cal Research Institute) mammalian proteins, and that this re- sults in the aggregation so often seen. Indeed, there has been one case in which bacteria engineered to overproduce their own chaperones successfully produced a mammalian protein that otherwise irretrievably aggregated. Unfortunately, this approach has failed in other cases. And no one has yet reported introduction of mammalian chaperones into bacteria to help produce soluble mammalian proteins. Yet this, along with the intro- duction of mutations that block the aggregation pathway and the discovery of small molecules that prevent aggregation, is one of the most promising ways to overcome the roadblocks that biotechnology companies have so often encountered. Partially folded intermedi- ates at the junction be- tween productive and off- pathway folding. General- ized pathways showing an inclusion body derived from an intermediate on the folding pathway. This illustration shows a speculative intermediate in the formation of an α/β protein in which a helical domain is docking against a sheet. In the inclusion body pathway, the same interaction proceeds between intermediates, resulting in a polymeric aggregate (3, 62). [Re- drawn from FASEB J. 10, 58 (1997)]

however, scientists often look for was that the mutant proteins experimental systems that will were less stable. After all, the allow them to get a foothold or temperature scale is fundamen- find a way around them. In this tally defined by how much atomic- case, they thought that a large scale shaking or motion is going protein, whose folding passes on; in other words, the higher the through multiple stages, would be temperature, the more shaking a good system for looking at fold- there is. This implies that a less ing pathways within cells. Many stable protein is more likely to fall temperature-sensitive mutations apart at elevated temperatures had already been isolated in and might therefore be more bacteriophages, but never exam- likely to end up (like cooked eggs) ined for their effect on folding. as insoluble gunk. Their hopes were realized: The But this turned out not to be majority of the temperature- the case. If the mutant chains sensitive mutations they found, were allowed to fold up at low despite having only one amino temperature, and were then acid altered, caused the tailspike heated, they were as stable as protein to end up as insoluble wild-type. It turned out to be a gunk at high temperatures. Since partially folded intermediate, on these folding failures were occur- the route from the random shoe- ring in bacterial cells that were lace to the correctly folded pro- growing in the laboratory, it was tein, that was sensitive to tem- now possible to analyze what perature. At higher temperatures went wrong in a protein’s folding these intermediates would stick to process. themselves and be unable to reach The obvious guess at the time the properly folded state. This turned out to be a general problem in the folding of many proteins: They have to pass through partially folded states in which they are delicately poised between folding all the way to the correct state or becoming seri- ously stuck as a result of prema- ture entanglement with other molecules. Recognizing that it was the intermediates and not the fully folded protein that were in trouble opened the way to under- standing some aspect of a range of diseases. Familial Amyloidotic Polyneuropathy Over the past several years, col- laborators have conducted similar studies in connection with a hu- man disease. The minor differ- ences between their results and others’ are very revealing. A ‘ribbon diagram’ showing two molecules of the protein transthyretin In the hereditary disease famil- docked together. The spiral coils of ribbon represent a-helix, while the flat ial amyloidotic polyneuropathy arrows running alongside each other represent b-sheet. (Generated with (FAP), peripheral nerves and Molscript by Scott Peterson, Texas A&M University) other organs are damaged by deposits of amyloid-type protein. under mildly acid conditions than Although the disorder is quite is the wild-type structure. This rare, extensive genetic studies contrasts with the P22 tailspike have shown that the disease mutations, which fold slowly but results from mutations in the are stable once folded. It also protein transthyretin. As with the appears that transthyretin aggre- P22 tailspike protein, gation takes place from a mono- transthyretin contains large meric unfolding intermediate, β amounts of -sheet structure and rather than the folding intermedi- normally consists of several iden- ate involved in P22 tailspike tical amino acid chains (four in aggregation (the pathway may or this case) associated into a single, may not be the same in both three-dimensional structure. directions). FAP results from any of more In both cases, however, the than 50 distinct mutations within single-chain intermediates have the transthyretin protein, each structures that nature has de- altering a single amino acid. signed for association with other After studying several of these, chains of the same type. It appar- scientists found that their four- ently takes only a very small chain structure is less stable change in the shape of these intermediates to alter their nor- Alzheimer patients. So the con- mal linkage with two or three sistent association of amyloid other chains into an endless series precursor protein mutations with of linkages that creates insoluble early-onset Alzheimer’s has gunk. finally answered a long-debated There is yet another contrast question: the deposition of neu- between the P22 tailspike muta- ritic plaque is part of the pathway tions and those in transthyretin: leading to the disease, not a late From the P22 virus’s view, the consequence of it. problem with the tailspike muta- To help understand the Aβ tions is that not enough normal aggregation process, researchers protein is made. People with chemically synthesized fragments transthyretin mutations, on the of the 40-amino-acid-long peptide. other hand, have all the normal By using these fragments, they transthyretin they need to carry showed that the key step is get- out its usual function (transport- ting started. Specifically, the ing the thyroid hormone). The precursor fragments have to form problem is that, as the protein is a specific nucleus, which then being broken down, it forms in- grows into the amyloid process. soluble gunk, and the insoluble Possibly the slowness of this first gunk poisons the tissues where it step is why Alzheimer’s disease is is deposited. almost entirely limited to older Alzheimer’s Disease people, and it could be that the mutations in amyloid precursor FAP is a rare disease; not so protein that lead to early-onset Alzheimer’s, which afflicts 10 Alzheimer’s are the ones that percent of those over 65 years old make it progress more quickly and perhaps half of those over 85. and easily. Every year Alzheimer’s not only Even so, Aβ remains soluble in kills 100,000 Americans, but also most people. Most individuals costs society $82.7 billion to care who develop Alzheimer’s disease for its victims. have the normal form of amyloid In 1991, several different re- precursor protein, indistinguish- search groups found that indi- able from that in people who viduals with specific mutations in never acquire the disorder. Why their amyloid precursor protein the same form of Aβ aggregates in developed Alzheimer’s disease as some people’s brain but not in early as age 40. The body pro- others’ remains a mystery, al- cesses amyloid precursor protein though a recent discovery has into a soluble peptide (small suggested an intriguing possibility. β protein) known as A ; under We know that people with β certain circumstances, A then different genetic variants of the aggregates into long filaments protein apolipoprotein E (apoE) that cannot be cleared by the have quite different risks of devel- body’s usual scavenger mecha- oping Alzheimer’s disease. Com- nisms. These aggregates then pared to those with the most β form the -amyloid, which make common variant, known as apoE3, up the neuritic plaque in those with the apoE4 variant are significantly more likely to de- velop the disease. Some studies suggest that those with the apoE2 variant may be at lower risk, although other studies disagree. These findings are particularly surprising because apoE is best known as part of the complex that transports cholesterol and other fatty materials in the blood- stream. What could a fat-trans- porting protein have to do with Alzheimer’s disease? It may be significant that small amounts of this protein are associated with Courtesy: National Institute on Aging, Bethesda, MD. neuritic plaque and that apoE binds to Aβ in the test tube. The results of this binding are in diseases transmitted by prions, or dispute, however. protein particles. Prions seem to Researchers report that adding be pure protein; they contain apoE to a test-tube solution of neither DNA nor RNA. Yet an soluble Aβ causes rapid formation infectious agent is necessarily of plaque-type β-amyloid fibers— self-replicating. How, scientists and that apoE4 does so more asked themselves, could a pure rapidly than apoE3. Others, protein replicate itself? however, have obtained opposite The answer now starting to results: apoE prevents fibril for- emerge may be viewed as a varia- mation. Thus, whereas some tion on the concept of the patho- suggest that apoE acts as a patho- logical chaperone, only in this logical chaperone, one that actu- case the protein serves as its own ally promotes misfolding, other chaperone. researchers believe that it exerts The protein whose aggregation a normal chaperone’s protective damages nerve cells in Mad Cow effect. In either case, apoE’s disease is constantly being pro- influence on the folding of Ab may duced by the body. Normally, play a major role in development though, it folds properly, remains of Alzheimer’s disease. soluble, and is disposed of without Mad Cow and Other Species problem. But suppose that some- how a small amount misfolds in a Perhaps the most interesting particular way so as to become a example of a protein folding disor- scrapie prion. If this scrapie prion der is Mad Cow disease and its bumps into a normal-folding human equivalent, Creutzfeldt- intermediate, it shifts the folding Jacob disease. These diseases, process in the scrapie direction along with the sheep version and the protein, despite its per- known as scrapie, have had the fectly normal amino acid se- scientific community in an uproar quence, ends up as more scrapie for years. They are infectious prion. And the process continues: So long as the body keeps producing Too Little, Too Late the normal protein, a little bit of scrapie prion can keep on creat- Despite the examples of FAP, ing more and then more. In Alzheimer’s disease, and Mad effect, the prion is “replicating” Cow disease, in which the prob- itself without needing any lem derives from accumulation of nucleic acid of its own. toxic, insoluble gunk, many hu- What old-school scientists find man diseases arise from protein even more strange is that the misfolding leaving too little of the process resembles something normal protein to do its job properly. akin to genetics. Different The most common hereditary strains of these diseases, with disease of this type is cystic fibrosis. somewhat different clinical Recent research has clearly symptoms, ‘breed true’ as they shown that the many, previously are transmitted from one animal mysterious symptoms of this or human to another. Moreover, disorder all derive from lack of a these strain differences are protein that regulates the trans- associated with slight differ- port of the chloride ion across the ences in the protein deposits cell membrane. More recently that apparently cause the dis- scientists have shown that by far ease. (Scientists have recently the most common mutation un- used these strain differences to derlying cystic fibrosis hinders the show that a few Britons truly dissociation of the transport- have Mad Cow disease, the form regulator protein from one of its seen in cattle, rather than the chaperones. Thus, the final steps usual human form of in normal folding cannot occur, Creutzfeldt-Jacob disease.) and normal amounts of active Just as replication can occur protein are not produced. without DNA or RNA, other A hereditary form of emphy- experiments have shown how sema shows an even greater ‘genetics’ is possible without analogy to the mutations studies nucleic acids. Thus, when re- in P22 tailspike protein. Investi- searchers mix seed quantities of gators have found that one of the two different scrapie prion most common mutations produc- strains in separate test tubes ing this disorder greatly slows the with large amounts of normal normal folding process, just as the protein, each test tube produces P22 temperature-sensitive muta- more of the specific scrapie prion tions do. As with the tailspike strain that was added. That is, mutations, the resulting buildup each strain induces the normal of a crucial folding intermediate protein to fold in exactly the same leads to aggregation, which de- way as the original seed. The prives affected individuals of α strain breeds true in the test enough circulating 1-antitrypsin tube, just as it does in the body. to protect their lungs. Emphysema Odd as it may seem, genetics with- is the result. out nucleic acid is truly possible in As intriguing as these examples the world of protein folding. may be, there is a far more com- mon instance of misfolding, which leaves too little normal protein to do its job. In this case, the protein’s job is to block cancer development. Over the past couple of decades, scientists have learned that most cancers result from mutation in the genes that regulate cell growth and cell division. The most common of these genes, involved in roughly 40% of all human cancers, is p53. The sole function of the p53 protein ap- pears to prevent cells with dam- aged DNA from dividing before the damage is repaired (or to induce them to destroy them- selves, if the damage cannot be fixed). In other words, p53 exists to prevent cells from becoming cancerous. p53 mutations associated with cancer fall into two classes. The Folding and aggregation during protein renaturation. first keeps the protein from bind- Correct folding reactions, leading to the native state [(1), ing to DNA; the other makes the (2)]. Irreversible aggregation reactions, starting from different conformations during the renaturation process [(3), folded form of the protein less (4)]. [From FASEB J. 10, 52 (1996)] stable. In the second group, there is simply never enough to a misfolded protein. Although properly folded protein around to many molecular biologists and block the division of DNA-dam- protein chemists believe this will aged cells. It will be interesting to be quite difficult, others are more see how many of the p53 mutants optimistic. fall into this second class and It is difficult to pinpoint where whether some way can be found to the search for treatment currently stabilize them. stands, however. One scientist Treating Protein Misfolding notes that the bulk of that work is tied up in the patent stage: com- The purpose of studying any panies are pursuing it but have human disease is to find ways to published little on the subject. treat it. The story of protein folding Nevertheless, one research group has not yet led to treatments for the has shown that both thyroid diseases involved, but this could hormone and the related com- happen within the next decade. pound TIP (2, 4, 6-triiodophenol) The key is to find a small mol- can stabilize transthyretin. Since ecule, a drug that can either stabi- TIP neither blocks the action of lize the normally folded structure thyroid hormone nor exerts any or disrupt the pathway that leads hormone-like effects of its own, it appears to be a promising treat- Suggested Readings ment for FAP. For a thorough discussion of protein Developing small-molecule folding, read a thematic issue on the therapies is quite straightforward topic in FASEB J. (1996) Protein Folding for proteins like transthyretin 10. that naturally bind small mol- A very short review of protein folding ecules, but these therapies are with superb illustration is: Jonathan more difficult to apply to proteins King (1993) The Unfolding Puzzle of that do not have a small-molecule Protein Folding, Technology Review, 58- binding site. 61. One of the few other groups For an overview of the role of protein currently publishing their re- folding in human disease see Gary search on small-molecule struc- Taubes’ (1996) Misfolding the Way to Disease, Science 271, 1493-1495. ture stabilizers is working to stabilize p53, an acknowledged For a short, comprehensive review of ‘difficult target’. In fact, one diseases due to protein misfolding, read Philip J. Thomas, Bao-He Qu, and Peter laboratory has obtained encourag- L. Pedersen (1995) Defective Protein ing results by using two different Folding as a Basis of Human Disease, approaches. TIBS, 20, 456-459. Treatments based on our grow- A good discussion of amyloid precursor ing knowledge and contined re- protein and its role in Alzheimer’s search of protein folding are on disease is: Celia Hooper (1991) An the way. When they arrive, the Exciting “If” in Alzheimer’s, The Journal saga that began with Pauling’s of NIH Research, 3 (April) 65-70. fundamental studies of protein For coverage of recent research on p53 structure and Anfinsen’s investi- and efforts to stabilize it, read: Rebecca gation of what some call ‘the L. Rawls (1997) Keeping Cancer in second genetic code’ will reach its Check with p53, Chemical & Engineer- ing News, 75 (February 18) 39- 41. practical fruition. For everything you might want to know about prion disease, read Richard Rhodes (1997) Deadly Feasts, Simon & Schuster, New York, and a review by Neil Stahl and Stanley B. Prusiner (1991) Prions and prion proteins, FASEB J. 5, 2799-2807.