Life's Greatest Secret

by Matthew Cobb

  *

  Nirenberg, who was at Berkeley, heard of the breakthrough over the telephone, and by 11 June he was back at Bethesda doing experiments. Matthaei recalled the feelings in the laboratory at the time: ‘of course we were excited, because we knew exactly what we had. And we knew what we’d wanted to get.’28 Everyone was sworn to secrecy – nobody was to hear about the finding until the results had been published. This caused some difficulties – at the beginning of June, Sydney Brenner gave a talk at Bethesda, during which he said it was not possible to study messenger RNA in a cell-free system. When Matthaei asked him how he could know this, Brenner astutely fired the question back at Matthaei, asking whether he had any insight into the question. Matthaei said nothing.29 Above all, the normally garrulous Tomkins had to bite his tongue throughout the Cold Spring Harbor meeting, which took place a week after Brenner’s visit. After the meeting was over, Tomkins finally cracked, and at the end of July he told Alex Rich in Boston. The news went no further, because Rich was too busy with his own work to gossip and, anyway, everyone else was either on holiday or was heading off to the International Biochemical Congress in Moscow.

  Matthaei had also found it hard to keep quiet. At the end of June he took the phage course at Cold Spring Harbor, during which each student described his or her research. Matthaei initially refused, but eventually outlined his discovery. Delbrück, who was teaching on the course, was amazed and immediately told Jerry Hurwitz at New York University; Hurwitz in turn phoned Tomkins and got confirmation.30 The secret was out, and by the beginning of August, researchers in Severo Ochoa’s laboratory in New York had heard the garbled news that ‘someone from MIT’ had broken the code.31

  Meanwhile, Nirenberg was heading for the Biochemical Congress in Moscow, where he was planning to unveil his discovery. Before he left he had two things to tie up. He got married to Perola Zaltzman, a Brazilian biochemist, and he staked his claim to priority by submitting two articles to Proceedings of the National Academy of Sciences. At the time, articles in PNAS had to be sponsored by a member of the Academy. Hearing that Academy member Leo Szilárd was staying just down the road in Washington, Nirenberg spent a whole afternoon discussing the results with him in the lobby of the Dupont Hotel. Szilárd was reluctant to help out: ‘It’s too much out of my field,’ he said, ‘I’m sorry, I can’t sponsor it.’32 It is hard to imagine Szilárd responding in the same way to Jacob or Monod. Nirenberg was clearly an outsider.

  The two papers were submitted to PNAS on 3 August 1961, with the support of Joseph Smadel, the Associate Director of NIH. Straight afterwards, Nirenberg flew to Moscow. The articles appeared, back to back, in the October issue of the journal, by which time everyone who was anyone already knew all about their stunning content. Both papers contained meticulous descriptions of the protocols involved and above all were characterised by the use of tightly conceived control experiments that enabled the authors to exclude alternative explanations, rendering their conclusions incontestable.

  The first, more technical, paper described the characteristics of protein synthesis in cell-free E. coli extracts, repeating and expanding the results that had been published earlier in the year. Significantly, Matthaei was the first author on this article, which was destined to be read by fewer people (it has been cited fewer than 300 times). This paper showed that protein synthesis could be disrupted by RNase, which attacks RNA, and – eventually and to a lesser extent – by the DNA-destroying enzyme DNase. Matthaei and Nirenberg correctly suggested that the presence of intact RNA was essential for protein synthesis to occur, and that inhibition by DNase was due to ‘the destruction of DNA and its resultant inability to serve as templates for the synthesis of template RNA.’33

  The second paper was obscurely entitled ‘The dependence of cell-free protein synthesis in E. coli upon naturally occurring or synthetic polyribonucleotides’. Despite this unappetising opening, it contained the experiment that showed an increase in the amount of radioactive protein when poly(U) was incubated with radioactive phenylalanine – after refining the protocol, they were able to get a roughly 1,000-fold increase over control levels. The article, which has been cited more than 1,400 times, was entirely couched in the language of biochemistry and protein synthesis, referring in a rather old-fashioned way to ‘specificity for phenylalanine incorporation’. Only in the final paragraph did Nirenberg and Matthaei frame their discovery in the new language of life, emphasising that the full detail of the code was not yet known:

  One or more uridylic acid residues therefore appear to be the code for phenylalanine. Whether the code is of the singlet, triplet, etc., type has not yet been determined. Polyuridylic acid seemingly functions as a synthetic template or messenger RNA, and this stable, cell-free E. coli system may well synthesize any protein corresponding to meaningful information contained in added RNA.34

  Most people assumed that the code was based on triplets, simply because this gave sixty-four possible combinations for twenty naturally occurring amino acids, but this had not been shown to be true. Nirenberg and Matthaei were rightly hedging their bets – strictly speaking, their data were compatible with the unlikely possibility that just a single base of U coded for phenylalanine. Although they again referred to messenger RNA, they did not cite any of the three recently published papers that had first used this term (the two Nature papers and the Jacob and Monod review in the Journal of Molecular Biology, all of which had appeared in May). Indeed, for reasons that remain obscure, Nirenberg never cited any of these articles.35 In a note added shortly before the article was printed, they included a recent result obtained by Matthaei while Nirenberg was in Moscow – poly(C) coded for proline. Two words of the genetic code had now been read, but it was still not clear how many letters each contained.*

  *

  The Fifth International Congress of Biochemistry took place in Moscow, from 10 to 16 August 1961. It was the largest conference ever held in the USSR – there were more than 5,000 participants, including 3,500 foreigners from fifty-eight countries – and was commemorated by a special Soviet postage stamp. With nearly 2,000 talks and up to eighteen parallel sessions, many of the presentations were poorly attended.36 Eight large symposia, including one organised by Max Perutz on ‘Biological Structure and Function at the Molecular Level’, were held in various Moscow University buildings.

  The congress opened in the Sports Palace of the Central Lenin Stadium on the outskirts of Moscow, where the slides were poorly projected and it was hard to see anything. One symposium had to be cut short due to a press conference that was held for the second man to orbit the Earth, 25-year-old Gherman Titov, who had returned to Earth on 7 August, after spending more than a day in space. Later, delegates gathered in a sunlit Red Square to see a parade to celebrate Titov’s return.37 This was at the height of the Cold War, and Russian superiority in space was extremely significant. Furthermore, while the congress was taking place, the Cold War got a bit hotter as the Berlin Wall began to be constructed on 13 August.

  Like every other non-plenary speaker at the massive meeting, Nirenberg was given a brief ten-minute slot to present his findings, which concentrated on the material from the second PNAS paper and concluded, after a last-minute edit, with the phrase he and Matthaei had used in their article: ‘One or more uridylic acid residues therefore appear to be the code for phenylalanine’.38 The tiny lecture theatre was partly filled with a big old-fashioned slide projector, and there were only a couple of dozen people in the audience.39 Watson later said that he ‘heard rumours that there might be an unexpected bombshell talk by Marshall Nirenberg’ – this may have been chatter from Delbrück or others, but according to Nirenberg he introduced himself to Watson shortly before the talk and outlined his findings.40 Whatever the case, Watson was clearly not intrigued enough to go and listen. Instead he sent along his postdoctoral researcher, Alfred Tissières; Matthew Meselson was also there. Meselson, who was younger than Nirenberg, later recalled:

  I heard the talk. And I wa
s bowled over by it. … I went and chased down Francis [Crick], and told him that he must have a private talk with this man.41

  The next morning, Watson told Jacob what Nirenberg and Matthaei had discovered. Jacob assumed that this was one of Watson’s tiresome practical jokes, and refused to believe him.42 Crick was sharper – on hearing the news from Meselson he immediately decided to invite Nirenberg to present his talk again the next day, in the symposium on molecular structure and function that Perutz had organised and which Crick was due to chair. In typically generous fashion, Crick was offering Nirenberg the opportunity to step into history as the man who had cracked the genetic code.

  According to Watson, Nirenberg’s plenary talk was ‘an extraordinary moment’. At the time, Crick reported that the audience was ‘startled’ by Nirenberg’s announcement – he later described it as ‘electrified’.* Even the modest Nirenberg recalled that the audience was ‘extraordinarily enthusiastic’.43 Meselson recalled that after Nirenberg’s second presentation ‘I ran up to Nirenberg and I embraced him. And congratulated him … It was all very dramatic.’44 Nirenberg was deeply touched by this gesture:

  The second time I gave the paper it was to a very large audience. The reception was really remarkable, fantastic. I remember Matt Meselson, who was sitting right up front. I didn’t know him at the time, but he was so overjoyed about hearing this stuff that he impulsively jumped up, grabbed my hand, and actually hugged me and congratulated me for doing that. I could have been part of a rock band or something! That meant an awful lot to me. It really meant more to me than all kinds of awards and what-not because it was genuine and spontaneous.45

  Meselson also recalled the effect that the talk had on the audience: ‘it gave some people who were in this field the immediate itch to get out of Moscow, to get back to the lab.’46 What they would do back in the lab was simple – they had to adopt Nirenberg’s technique. Jerry Hurwitz, who heard both of Nirenberg’s Moscow talks, recently told me how the new approach changed everything:

  I remember thinking about the ramifications of the Nirenberg–Matthaei findings. In early June 1961, at the Cold Spring Harbor meeting, it was evident that a number of laboratories were using specific proteins … to get at the code. I recall thinking that these efforts were now obsolete.47

  Harold Varmus, who was not even a scientist at the time, found his life changed because of Nirenberg’s talk. Varmus, a student of English, was accompanying his biochemist friend Art Landy, who was attending the congress. Varmus understandably spent the day that Nirenberg spoke ‘riding Moscow’s fabled ornate subways and roaming Russian art galleries’. But that evening he heard something that made him doubt his career choice:

  listening to Art Landy’s excited report at the end of the day in our rooms at Moscow State University, I began to understand that something of fundamental significance had occurred, and I felt that a seed of professional envy had been planted. Scientists seemed likely to discover new, deep, and useful things about the world, and other scientists would be excited about these discoveries and eager to build on them.48

  That feeling grew, and Varmus soon switched from English to medicine. In 1989, he won the Nobel Prize in Physiology or Medicine for his work on genes and cancer.

  Not everyone was convinced, however. When Jerry Hurwitz returned to New York University in the middle of August, he told colleagues about Nirenberg’s talk but added:

  several people didn’t believe the validity of the data. It seems that, although Nirenberg performed the first basic experiment, there is still a great deal to gain from the application of this new, extremely sensitive method.49

  Whoever those ‘several people’ were, within weeks their doubts were blown away.50

  Matthew Meselson later explained the widespread surprise that was felt about Nirenberg’s success, in terms of the social dynamics of science:

  there is a terrible snobbery that either a person who’s speaking is someone who’s in the club and you know him, or else his results are unlikely to be correct. And here was some guy named Marshall Nirenberg; his results were unlikely to be correct, because he wasn’t in the club. And nobody bothered to be there to hear him.51

  This explanation is reinforced by a private letter to Crick, written in November 1961 by the Nobel laureate Fritz Lipmann, which celebrated the impact of Nirenberg’s discovery but nevertheless referred to him as ‘this fellow Nirenberg’.52 In October 1961, Alex Rich wrote to Crick praising Nirenberg’s contribution but wondering, quite legitimately, ‘why it took the last year or two for anyone to try the experiment, since it was reasonably obvious’.53 Jacob later claimed that the Paris group had thought about it but only as a joke – ‘we were absolutely convinced that nothing would have come from that’, he said – presumably because Crick’s theory of a commaless code showed that a monotonous polynucleotide signal was meaningless.54 Brenner was frank: ‘It didn’t occur to us to use synthetic polymers.’55 Nirenberg and Matthaei had seen something that the main participants in the race to crack the genetic code had been unable to imagine. Some later responses were less generous: Gunther Stent of the phage group implied to generations of students who read his textbook that the whole thing had happened more or less by accident, while others confounded the various phases of Matthaei and Nirenberg’s work and suggested that the poly(U) had been added as a negative control, which was not expected to work.56

  In fact, Nirenberg and Matthaei were not the only ones to have the idea of using synthetic polynucleotides to crack the code.* It had occurred, separately, to at least two people in Severo Ochoa’s laboratory. In 1957, the Yugoslavian scientist Mirko Beljanski was on sabbatical in Ochoa’s lab. For nearly a year, Beljanski tried to get synthetic poly(A) to direct protein synthesis in a cell-free system, but without success.57 Paul Zamecnik later recalled that Ochoa had sent him some poly(A) to be studied in the cell-free system, which was waiting in the freezer when the news of Nirenberg and Matthaei’s discovery came through.58 It is not obvious what would have happened had Zamecnik tried the poly(A); in 1961, Nirenberg and Matthaei, like Beljanski, failed to get poly(A) to produce any protein. The eventual explanation was quite simple – the protein produced by poly(A), a lysine polymer, interacted with some of the chemicals that were initially used in the cell-free system and gave no apparent result.59 A biochemical tweak was needed to make poly(A) work, but that would not have been apparent without the previous success of the poly(U) experiment.

  At exactly the same time as Nirenberg and Matthaei were tying up their experiments in the summer of 1961, Peter Lengyel, a young researcher in Ochoa’s laboratory, came up with the idea of using synthetic polynucleotides while listening to Sydney Brenner’s talk about messenger RNA at the Cold Spring Harbor meeting. While Ochoa was on holiday in Europe, Lengyel and three young colleagues planned out the experiments, but these were radically changed at the beginning of August when they heard through the grapevine that Nirenberg and Matthaei had used poly(U) to produce polyphenylalanine. The New York group immediately replicated the result. So at the same time as Nirenberg was stunning the audience in Moscow, Lengyel was showing that, in his hands too, poly(U) led to the incorporation of radioactive phenylalanine into a protein. By the time Ochoa returned to New York at the beginning of September, his laboratory was ready to make up the ground they had lost to the upstarts from Bethesda.

  *

  In late September, Nirenberg presented his work at a meeting in New York. He had barely progressed since Moscow – in an uncharacteristically unfocused moment, he had spent his time in the laboratory tying up a minor loose end in the protein synthesis pathway that had been activated in the poly(U) tube. He had even allowed himself a ‘leisurely, two week vacation’ in Copenhagen with his new wife.60 Nirenberg was therefore devastated when Ochoa stood up at the New York meeting and presented data that showed that the Ochoa group was hard on his heels. They had made huge progress – in the space of about six weeks, Ochoa’s lab had managed to get the cell-free syste
m working, had replicated Nirenberg and Matthaei’s results and above all had shown that two other artificial polynucleotides – poly(UA) and poly(UC) – were also active. As Ochoa later explained: ‘When we heard the news from Moscow we immediately tried it. And other polymers, copolymers we had in the icebox. We got immediate results with four or five.’61 New words in the genetic code were being read, but not in Nirenberg’s lab. Ochoa reported that poly(UC) led to the incorporation of radioactive phenylalanine, serine and leucine into a protein, and poly(UA) led to the incorporation of phenylalanine and tyrosine. Ochoa later described the excitement of discovery:

  Lengyel, Speyer, and I were watching the counter and were thrilled. This result, obtained for the first time anywhere, showed that the incubation of E. coli extracts with copolynucleotides containing C or A besides U residues promoted the synthesis of polypeptides containing serine, leucine and tyrosine, along with phenylalanine. I remember this as one of the most exciting moments of my life.62

  Understandably, Nirenberg’s response was less enthusiastic: ‘It floored me, that Ochoa had made such advances. Things developed much faster than I would ever have dreamed that they’d develop.’63 As he later recalled:

  I flew back to Washington feeling very depressed because, although I had taken only two weeks to show that aminoacyl-tRNA is an intermediate in protein synthesis, I should have spent the time focusing on the more important problem of the genetic code. Clearly, I had to either compete with the Ochoa laboratory or stop working on the problem’.64

  A colleague remembered the scene: