When a new protein slides off the tiny molecular assembly line within the cell, it is nothing more than a droopy string of amino acids, not yet fit for its designated profession. Only after being spun and pleated and braided into its proper three-dimensional conformation will a protein burst to life, seizing up oxygen if it is hemoglobin, shearing apart sugars if it is an enzyme, or lashing cells together if it is the stout twine of collagen.

Until recently science had scant idea how a simple chemical strand manages to fold into a working protein, with all its knobs, clefts, sheets, and curves arrayed in vivid harmony, able to mingle with other molecules around it. The problem is no mere academic exercise, but a question of central importance to biology. Proteins perform most of the tens of thousands of tasks needed to keep the body alive, and only a perfectly folded protein is up to the demands made upon it. Protein folding was thought to happen spontaneously, a newborn protein, or polypeptide, springing into its correct three-dimensional shape on its own, driven solely by the repellent or attractive electrical and chemical forces of its individual amino acids. As scientists tried to calculate what those enormously complicated interactions might be, they made little headway in cracking the folding conundrum.

So it is a great surprise to discover that nature doesn't let proteins fold up by themselves but has created a whole family of proteins whose sole purpose is to help other proteins crinkle and furrow. The detection of the handmaiden proteins, called chaperones, means that the traditional theory of spontaneous folding is mistaken: the forces inherent in a polypeptide's se- quence of amino acids aren't enough to sculpture and knead a protein into its correct, muscular form.

Instead, protein folding, like so much of what happens in the body, turns out to be done by committee. As the amino acid chain rolls forth from its birthing chamber in the cell, folding must begin immediately, and to that end, successive bevies of chaperone midwives rush over and gently embrace the flat polypeptide at hundreds of key spots, shielding it against the hostile environ- ment of the cell. The chaperones allow those amino acids that are destined for the interior of the active protein to curl in on themselves, while they encourage those regions meant for the exterior to turn and face the outside world. They help twist some stretches into corkscrews and pummel others into flat sheets. The chaperones also protect the fragile chain from becoming en- snarled with other infant peptides floating in the cell, as it would if left unattended.

Nor do the jobs of chaperones end once the initial folding is through. Should the cell suffer a shock from extreme heat, oxygen cutoff, or any sort of trauma that threatens the structural integrity of the thousands of proteins within, the chaperones will toil mightily to prevent protein disintegration, latching on to the wilting molecules and helping to bend them back into shape. So indispensable are the folding molecules to growth and survival that cells experimentally deprived of their chaperones rapidly die.

The information that researchers are gathering about chaperones could yield knowledge about many of our worst afflictions. After a heart attack, for example, large patches of cardiac muscle atrophy and die as a result of temporary oxygen deprivation. If the chaperones in heart tissue could be manipulated right after a heart attack, the restorative molecules could shore up the collapsing proteins in the heart cells and perhaps prevent tissue death. Certain genetic disorders, like muscle-wasting diseases, may result from mutations that slightly weaken the cell's chaperones, leaving many proteins in disarray. In theory, a better understanding of chaperones will improve drug design. Many drugs currently in use and under investigation are based on natural proteins. If pharmaceutical companies could learn why a par- ticular amino acid prompts a protein to twirl up rather than down, and why the protein works better in one shape than in another, they could mix and match components to improve on nature's offerings.

Before we get swept away with easy enthusiasm, though, we should bear in mind that chaperones are only a part of the protein-folding puzzle; much remains to be learned about the wildly complex dynamics of protein structure. The information needed to determine the final working profile of the protein is inscribed in the sequence of its amino acids, and scientists still do not understand the electrochemical pushes and pulls of those building blocks. Chaperones do not dictate how a protein folds but only help the protein realize its ambitions and steer it away from binding with bad company. In other words, they're the foremen of the floor: they make sure the job gets done right, but they're not the ones who decide what the final product should look like. Those specs, written in the amino acid code of the protein, have yet to be deciphered.

Nevertheless, the discovery of chaperones is a bright spot in an otherwise discouraging discipline. By tracing the interactions between a folding protein and its industrious assistants, scientists just may be able to identify all the intermediate steps between a flaccid polypeptide and a strapping folded protein.

The reason that scientists so long neglected to recognize chaperones is that they carried out their studies of protein folding in vitro, throwing together isolated protein subunits and a few other ingredients to see what emerged. Biochemists found that nearly any polypeptide they tossed into the test tube would, if cultivated under exacting conditions, fold into its active shape as the different amino acids on the chain looped in one direction or another, depending on their inherent molecular properties. This discovery led researchers to assume that protein folding happened spontaneously in the cell as well. In an effort to understand folding dynamics, they used intricate mathematics, the principles of physics, computer graphics, and difficult crystallography techniques, but with only modest progress.

In the mid-I98os, scientists who worked not with isolated molecules but with living cells detected several proteins that flared into action when the cells were subjected to abnormally high temperatures. Those responders, named heat-shock pro- teins, were determined to playa crucial role in helping the cell weather the heat by stabilizing all the rest of the proteins, which otherwise would unravel.

Biologists, after identifying many kindred of the original salvation proteins, divided them into at least two superfamilies of proteins and observed them in creatures across the evolutionary landscape, from bacteria to humans. A big advance occurred when those same heat-shock proteins were spied in normal cells that had not been stuck into a laboratory oven, suggesting that stress proteins participated in the daily life of the cell and did not serve only as an emergency crew.

Geneticists then discovered that yeast cells harboring mutations in their heat-shock genes were in terrible shape, a mess from their nuclear heart to their rubbery membranes. The proteins in these cells hadn't folded properly, a defect that led to wholesale havoc. The biologists realized they had an unexpected bonanza: a class of proteins to cast light on the blackness of the folding problem. That is when the proteins were rechristened chaperones to reflect their more general duties, although they are also called stress proteins, GroEI, or any number of unevocative names.

Most of the chaperone experiments to date have been performed in yeast or bacterial cells, which are easily manipulated, or in isolated cellular structures, like the mitochondria, where the body's energy is produced and where many proteins must be created and folded to help stoke the power plant. Researchers now understand that, from the perspective of a newborn poly- peptide, conditions in a living cell differ dramatically from those in a test tube, and that chaperones serve as indispensable nurse-maids. One type of chaperone after another steps in to assist as folding begins, a process that takes an average of three or four minutes. "It's like Snow White and her seven dwarfs," Dr. Mary-Jane Gething told me. "One dwarf has the hammer, another the chisel, a third the shovel, and so forth."

As a stretch of amino acids begins rolling off one of the cell's ribosomes, the pear-shaped factories where proteins are created, a small chaperone called hsp7O drifts over and grips certain tender areas of the polypeptide. The chaperone recognizes stretches of amino acids that are hydrophobic, or water-hating. Such hydro- phobic patches are destined to end up tucked inside the protein once folding is finished. But until they can curl under, they are vulnerable to misguided merging with other polypeptides, and thus must be safeguarded by the chaperones. The concentration of proteins in the cell is as thick as honey, and young proteins must be sequestered from the ambient ooze. During the early stages of folding, the polypeptide may form characteristic cork- screw shapes, or linked loops that resemble a Christmas bow, or slender fingerlike projections. When the preliminary folding is complete, members of the first round of chaperones relax their grip and drift off.

As folding proceeds, and the bows and corkscrews really begin to twist in on themselves, a second group of chaperones, called hsp6o, takes over. This molecule looks like two doughnuts stacked on top of each other, offering a snug tunnel into which the partly folded protein is pulled for further twisting without interference from stray peptide strands outside. Eventually, the protein torques into something round, dense, and energetic, perhaps sporting a pincerlike cap to snare hormones or a deep groove to capture a foreign microbe. Upon the completion of folding, the chaperones let the protein loose to try its luck in the thick of life and move on to the next newborn in need of care, approaching and abandoning the whorls and loops and jags of an ever-tightening peptide chain. Hundreds if not thousands of times each hour, they are alchemists, spinning dull chemical straw into a splash of protein gold.