Paul Lauterbur and the Invention of MRI

by M. Joan Dawson

  Paul had always been interested in polymers; it was for polymer research that he had embarked on NMR in the early 1950s. But he was then thinking of rather simple polymers. Proteins are the most complex and varied of polymers, large biological molecules made up of twenty-three different amino acids and occurring in strands up to tens of thousands of amino acids long. They fold and twist into three-dimensional structures, adding both to the complexity and to the usefulness of succeeding to obtain their NMR spectra. A major drive to Paul’s study of biological polymers in 1960s was to show others that it could be done.

  During the 1960s, serious proton NMR spectroscopy of proteins was being carried out, but because the protons’ signals are crowded together, information was difficult to extract. The 1H spectra of proteins looked like lumpy mountains, and interpretation ranged from very difficult to impossible. Until the early 1970s, studies of proteins were generally limited to certain signals that either fortuitously appeared outside the broad unresolved hump or were made to by disruption of the three-dimensional structure, which is often the part you are interested in. Attempts were made to see whether chemical shifts could be utilized to determine the order in which amino acids are joined to make a protein. The effects were small, however, and other techniques proved much better for carrying out sequence analysis.

  Paul recognized that 13C NMR has unique and powerful advantages over its proton NMR cousin. In 13C spectra, the signals are widely separated so the “lumpy mountain effect” that plagued 1H spectroscopy does not exist. The issue in 13C NMR is not one of dispersion, as in 1H spectroscopy, but of signal strength, because of the weak intrinsic signal intensity of 13C and its low natural abundance. Paul was exceptionally good at pulling out 13C signals.

  Sabbatical: Sunshine and Shadow

  In reorienting his research toward problems in chemical biology, Paul took the only sabbatical of his life during the academic year 1969–70. Scientists usually use the sabbatical year for travel to another university to learn new techniques and to interact with people doing work related to their own. In Paul’s case, his exploration of new directions with an exciting group of colleagues would have a decided impact on his later research and on the outside world. He would be with John Baldeschwieler’s group in the Chemistry Department at Stanford. The sabbatical would be highly productive, the antithesis of the extended vacation in exotic places imagined by cynics. He would work with Tom Link in George Stark’s laboratory in the Stanford’s Medical Center on 13C labeling of proteins, and with the pharmaceutical company Syntex on tritium (3H, the radioactive form of hydrogen) NMR. He also linked up with Varian Associates and had the opportunity to work with their new superconducting magnet-based NMR system. No one, not even the inventors, knew that superconducting magnets, which could increase field strengths by orders of magnitude, were about to explode into NMR history.

  So he packed his car with the two children, Dan and Sharyn, ages eight and six, rented their Long Island house to people he hoped (in vain) would take good care of it, rented for himself another house in Palo Alto, and drove with Rose Mary across the country. They took nearly a month zig-zagging up and down the country as Rose Mary and Paul showed their children their homeland of America.

  The children were astonished by the nearly perpetual summer in California. They could hardly believe they were roller-skating in t-shirts on Christmas day instead of bundled up in parkas to make snowmen. Rose Mary enjoyed it, too. She was involved in many activities. Her husband and children were happy. The weather was delightfully warm. She was irritated that Paul made frequent trips back to Stony Brook to meet with students, but he left her with two cars, a map, and relative freedom. She was involved in the local women’s movement. She had read Betty Friedan’s The Feminine Mystique, the book that launched the women’s liberation movement, when it first came out in 1965. Rose Mary threw herself into the cause with all the enthusiasm of a true believer. Paul felt that some of her activities were directed personally at him. She once kept a placard in the kitchen that read, “Down with men!” As usual, Paul was mostly in the laboratory, and even when he was at home he was working.

  Paul had left undergraduate Jerry Ackerman behind to simulate mathematically 13C spectra of proteins, and two graduate students, Jose Ramirez and Skip Hutton, studying isotope effects on NMR spectra. He flew back to Stony Brook almost once a month to stay in touch with these activities. Ackerman was perfecting a computerized system for summing spectra to increase the size of the signal, and the group soon succeeded in obtaining partial 13C NMR spectra (the C=O region) of lysozyme in a standard magnet. (Lysozyme is one of the smaller proteins, so the results are relatively easily interpreted. It is used a great deal in basic research.) This was a huge effort, since they had to improvise a modern experiment using outdated equipment. Had the lab been more blessed with resources they could simply have bought a computer-averaged transient box from Varian. This was neither the first nor the last time Paul and his lab members succeeded in doing state-of-the-art experiments with ancient equipment. They were always very clever with this, but at great cost in lost research time.

  Jerry’s study was noteworthy. He and other students simulated the 13C spectra of all the amino acids in lysozyme, assuming different line widths and coupling constants for the signal at different field strengths. They then added together the amino acid spectra to see what the spectrum of a denatured protein would look like. (A denatured protein has all of its chemical bonds intact, but the three-dimensional structure is destroyed, and it does not work.) Jerry and Paul also modeled mathematically the spectra of the native protein with its three-dimensional structure intact. To do this, Paul chose to simulate the results expected “at an impossibly high field strength of 300 MHz,” when 60 MHz was the norm. Higher field strengths were a fantasy, said Paul, as in, “Why don’t we get one for phosphorus at 600 MHz? That would be nice.” This was in the category of “and pigs may fly.”

  Could useful 13C signals be obtained from entire proteins by actual experiments? Part of Paul’s research plan for his sabbatical was to show that 13C NMR spectroscopy could prove useful in studies of function and structure of macromolecules. It had now been a dozen years since he produced the first 13C NMR spectra of small molecules, and he wanted to push the envelope even more. “I was the only one foolish enough to think it could work with big molecules in dilute solution,” he said of this period. A successful experiment was carried out with Leroy Johnson on an ordinary protein using only the naturally occurring 13C.6 Although Paul said those first spectra weren’t worth a whole lot, they opened up a research area that was to become very fruitful in later years.

  Figure 5.4

  Paul at his desk at Stony Brook.

  His more ambitious project was to enrich the groups at the protein’s active site with 13C, in order to get larger signals from this region. This effort was a study of Murphy’s law in research.7 In their first attempt, the cooling system failed, and the sample was cooked to a yellow gunk. The spectrum did show four broad lines, but they were uninterpretable. A noisy but tantalizing spectrum was obtained in the second attempt, after more than three and a half days of data collection. Paul then tried a different synthetic approach.8 He bought 13C-labeled glycine from Bio-Rad Corporation as the starting material. But when he checked it for purity he found he had been sold not glycine at all but ammonium chloride, a by-product of glycine synthesis. The people at Bio-Rad, not understanding NMR technology and not knowing that the spectra Paul sent them were incontrovertible proof of their error, quite ridiculously insisted there could be no problem with their glycine! By this time Paul’s sabbatical was ending, and there was no way to recover his project.9

  Paul was finally able to publish the first isotopically enriched 13C NMR spectrum of a protein, hen egg white lysozyme, in 1970.10 But his methods were now outdated. Paul noted in this paper that they had reached about the highest sensitivity possible using standard continuous-wave technology. The future for the routine study of e
nzymes lay in the new pulse and Fourier transform technique, which in effect gave much more visible NMR signals. This revolutionary method had recently been introduced by Wes Anderson and his then postdoctoral fellow, Richard Ernst (who would win the Nobel Prize in Chemistry in 1991). Richard had joined Wes at Varian Associates, and together they worked out the new methodology—in Varian’s laboratory, on Varian instruments, and with Varian’s money.11 Wes “had a trained Swiss monkey at the time” laughs Richard.

  Losses at Varian

  Since pulse and Fourier transform NMR was to become the most widely used of NMR techniques, and since Varian brought it into being, Varian should have patented the technology and pushed hard for its acceptance. Russell Varian had patented just about all NMR technology and assigned the patents to Varian Associates even before he had formally set up the company. Earlier, Felix Bloch had patented nuclear induction at Russell Varian’s request. It gave Varian a monopoly for several years. Varian collaborated with scientists at Stanford to commercialize quickly the continuing advances from that group. Jim Schoolery ran the Varian applications laboratory (the Aplab) with a program to make NMR spectroscopy a standard tool in analytical chemistry. The great A-60 was produced by Varian. “Varian” meant NMR in the same way that “Hoover” once meant vacuum cleaner.

  But by the late 1960s, “Varian had a messed-up culture that even when hit three times with a 2 × 4 didn’t get the message,” Paul told me. “The Varian powers decided Fourier transform was too difficult a technique and not necessary anyway, since they owned the market in NMR spectrometers.” As late as 1969 Varian put out a new-generation XL-100 spectrometer as the same old low-sensitivity continuous-wave instrument, configured in such a way that made the eventual addition of Fourier transform hardware cumbersome. Varian made another bad decision at about this time. “Jim Hyde was the EPR man for Varian. They supported him. Then the big wigs said, ‘EPR is a mature technology.’” Hyde, apparently no longer useful to Varian, went to the Medical College of Wisconsin, where the university built him a laboratory and he did good work in EPR and EPR imaging.

  Bruker Instruments, led by Gunther Laukien, now saw both opportunities, and jumped in. “They made EPR machines and they got 100% of the market.” Some scientific meetings can be a little dull, but now and then one comes along that wakes everyone up. The Disneyland meeting (in Anaheim, California) of the Society for Applied Spectroscopy in 1969 was one of those. “FT,” Paul said, “really happened at the Disneyland meeting, where Bruker first showed a lot of results from their [the first] commercial FT machine, in their hospitality suite at the Disneyland Hotel.” Everyone was impressed. Paul said, “People were astounded at the sharp 13C lines in their spectra. The Varian technical people were practically sobbing in the corner. That was Bruker’s big breakthrough.”

  With superconducting magnets and pulse and Fourier transform methods, routine 13C spectroscopy ripened. Paul had long predicted not the pulse and Fourier transform method and not superconducting magnets, but that a way would be invented to make 13C NMR work more easily and more routinely.13C NMR spectroscopy was a particular beneficiary of the new technologies, which so greatly amplified the tiny signals; there would now be an explosion of 13C NMR applications to biochemical systems that still hasn’t stopped.

  Kivatec, the Hole to Nowhere

  In another effort during his year at Stanford, Paul, his host John Baldeschwieler, and a previous visitor to the Baldeschwieler lab, Paul’s old friend Barry Shapiro, together with B. B. McIntyre founded a company to separate 13C in much larger quantities than were then available and to synthesize its compounds. McIntyre was an employee of the National Research Laboratory at Los Alamos at the time. The idea was to use the technique perfected at Los Alamos of distilling 13C, using distillation columns dug deep down into soft rock on the plateau of Los Alamos. The name they gave their company, Kivatec, was meant to commemorate the kivas of the Indians of the Southwest, dark, mysterious places dug deep in the earth. There was a complication that 13C was being enriched commercially at the time, for example by the Mound Laboratory, managed by Monsanto for the Atomic Energy Commission, which used a much more expensive method to harvest the substance, so very few chemical or biological scientists could afford it.12 Much valuable work in these fields was not being done. Los Alamos produced 13C at far less cost.

  The men participating in Kivatec felt they were doing good work for science.13 “Technology transfer” from government labs to commercial use was an important buzz phrase at the time; the dollars to support the research had ultimately come from taxpayers, and they, both scientists and politicians have often reasoned, should reap the full benefit. The founders of Kivatec also hoped they could make a little money. On the other hand, the National Laboratories are not allowed to compete with private industry. Because Monsanto made the stuff, the National Research Laboratory at Los Alamos could be seen as being in cahoots with its competitor. There is a delicate line between technology transfer (good) and competing with commercial interests (bad), a problem to be solved by the National Research Laboratory’s doing the rather undefined “fair thing.”

  The Kivatec effort eventually died because B. B. McIntyre’s bosses didn’t approve the plan. But its founders had some good times. They joked about feeding a 13C-enriched diet to mice and then feeding the mice to football players to make them heavier. At a Gordon Conference, where invited scientists get plenty of time to sit around and talk, Paul, Oleg Jardetzky, and Bob Shulman decided jokingly it should be a national priority to feed 13C to whales, so they could harvest enough 13C-enriched myoglobin to do a useful study of this important oxygen reservoir.

  Au Revoir

  His year in Palo Alto had been one of the most exciting of Paul’s life. Freed of other duties, he concentrated on his research, which was going through a very productive period. Rose Mary’s illness was in remission most of the year, and the children were happy. Though the more ambitious projects with 13C- and 3H-labeled proteins were still incomplete at the end of the sabbatical year, Paul was “well on my way to developing a program for using isotopic labeling to study proteins, and getting excited about the other opportunities in biology.”

  Coming home was a rude awakening. He returned to a house trashed by the irresponsible renters and found the same arguments continuing in his old department that had been going on when he left. Saturday nights at an off-campus pub, a tradition that had begun years before, were still in session. “We would sit and talk for hours,” recalled Thomas Irvine, professor in the Department of Mechanical Engineering.14 Paul resumed his place and continued with these sessions until he left Stony Brook.

  Rose Mary’s illness had begun to creep back toward the end of the sabbatical year, in the spring of 1970. She was lonely and missed her friends. She felt that Paul was less patient and angrier about her illness than before. Nineteen seventy-one, the year after the sabbatical, she says, was the worst year of her life. “I would call Paul in his office, in tears, it was so bad,” she recalls. She suffered nine months of pure depression. She thought of suicide every day, but didn’t know how to do it. She really wanted to live, but what was happening then was not living. In September she returned to North Nassau Mental Health Center, hoping to be evaluated as a schizophrenic, because there were drugs for this illness. But she flunked the test. They did thirteen more shock treatments and put her on a regimen of megavitamins and a hypoglycemic diet. She lost weight, and the treatments seemed to help. For almost a year she was convinced she was going to be all right.

  Baby Enron

  It was really more like the pop of bubble wrap than the atomic explosion of Enron, but it had many of the same characteristics. In one short year Paul’s life went from high to low. Spring and summer of 1971 were a nadir in Paul’s life. Rose Mary’s illness was particularly severe, and he was unable to attract grant funding for his research. He was involved with a troubled company that would smear his name and make him enemies. He called NMR Specialties “the double res
onance disaster.” (The company had been started to make double resonance equipment.) NMR Specialties was, in fact, a very special company that thrived on producing and selling unique components and accessories for NMR experiments. In addition to their own productions, they found and packaged for simple NMR experiments unique items that were not easily available elsewhere. They looked for products and markets that were not supplied by the bigger guns. Many scientists found the functions of NMR Specialties nearly indispensable to their work. Experts in the field could make their own products, but most chemists relied on NMR Specialties to make their work easier. They supplied the donuts for the ENC. What could be more upright and responsible?

  This is the story as Paul told it to me. A decade earlier, Paul Yajko, a Varian service engineer with whom Paul had collaborated, planned his own start-up and convinced several prominent people to become members of the board, Paul among them. His founding plan was to make a device that could increase the visibility of NMR signals by suppressing the cross talk between different nuclei. Varian had made such a spin decoupler, and although the first one did not work well, people saw its possibilities and clamored for it. The trouble was that each scientist wanted something a little different, so although Varian had built the prototypes, the company decided it couldn’t accommodate all the conflicting demands and that the decoupler could not be commercialized profitably.

  Yajko planned to copy the discarded Varian design and build a small company, NMR Specialties, to manufacture it. He had a ready market for his machine among the scientists, including Paul, who had gotten excited by the new technology that had suddenly been dropped. Another product they started making early was a pulsed spectrometer apparatus for making T1 measurements, an instrument that worked well. While Yajko’s spin decoupler was serviceable in some ways, his customers loudly pointed out that it could have been much better. Yajko neither knew himself how to improve it nor allowed his engineers to spend time making it better. Looking back, Paul thought this was an early sign of bad things to come. Perhaps another sign that a more skeptical man than Paul would have attended to was the business card of Herman J. Israel, the banker who supplied the start-up loans for NMR Specialties. He identified himself as “The Jewish Godfather,” and added a line at the bottom: “I am a reasonable man.”