ERWIN CHARGAFF 350 Central Park West, New York, New York 10025

Annals of the New York Academy of Sciences (1979) 325, 345-360.[With the permission of Erwin Chargaff]

[Comments by DRF in square brackets.]

NOWADAYS WE ARE, of course, familiar with such terms as "molecular biology," "molecular genetics" or "molecular pharmacology." The curious umbrella of molecularity under which the various biological disciplines practice a previously unimaginable form of togetherness testifies to the extent to which chemistry -- the very science of molecules and, therefore, one of the few not tolerating the all-embracing adjective -- has acted as a cement holding together the several branches of biology. In this respect biochemistry naturally shares the role of chemistry, and nobody has yet come forward with such a thing as molecular biochemistry. If we had to give up molecules, what would be left of us? Well, perhaps molecular biology would be left.

    There were, however, times when the biological sciences were not yet domesticated and did not march nicely in pairs on the leash of chemistry. They were robust fellows, each with his own code of honor and jealous of his independence; and they left each other more or less alone. I remember these days quite vividly, and how astonished we students were when at the chemistry colloquium at the University of Vienna a stray botanist or pathologist put in an appearance. In an even earlier generation the great Karl Landsteiner was one of the exceptions. When I met him in 1935 in Siasconset on Nantucket Island and he told me, walking on the beach or sitting in his somber house, of his early days in Emil Fischer's laboratory in Berlin, I was surprised about the wide range of his scientific interests. Now, when I am much older than he was at that time, I realize that this form of openness is no longer possible, and the sciences, as they have grown together, have become more hermetic than they ever were. There existed, of course, and there still exist, a few people able to break through the boundaries.

    Of the exact sciences, physics and somewhat later chemistry were the first to develop greatly. Before they reached the stage at which they could support the biological sciences, it is not surprising that biologists had little use for chemistry and physics. For this reason, biochemistry, not to speak of biophysics, represents a relatively late development. As was to be expected, among the biological disciplines, physiology was perhaps the first to experience a need for chemical assistance; and here lies, in fact, the origin of biochemistry. It began as a branch of physiology and was, for a long time, designated physiological chemistry.

    The "molecular revolution" --10% advance and 90% verbiage -- came about, however, in a surprisingly different fashion. Not so much physiology as two other branches of biology, microbiology to a great extent and immunology to a lesser, were involved; and the real beneficiary was a fourth branch, namely, genetics. Of immunology I do not want to speak here. I already have mentioned one of the great men who brought about the association of immunology and chemistry, Karl Landsteiner; and it is moving and surprising to me that many of us know the appealing figure of Michael Heidelberger, the true founder of what is essentially a science in itself, immunochemistry.

    I should, however, like to say a few words about genetics, doing justice to the title of my talk. If we are to believe the Oxford English Dictionary, the word genetics was first used in 1905 or 1906 [Huxley used it at least as early as 1862]. The man usually credited with the introduction of this term is the English biologist William Bateson. It is one of the youngest sciences, being dated from the year 1900, when Gregor Mendel's observations were rediscovered. To the extent that its first tools consisted in breeding experiments, genetics also is a very old science indeed: animal and plant breeders practiced an early form of applied genetics. For some reason popular ideas of heredity have always involved some kind of "blood and soil" mythology. Therefore, I could read only recently in a book on the Bach family that some of Johann Sebastian Bach's blood still rolls through the vessels of a Mr. Colson [This was near the time of the Watergate investigation in the USA.]. The author would clearly have been unable to tell me, more scientifically, what percentage of J S. Bach's DNA still was around. Even if he could have done so, I daresay it would have been of no interest whatever. DNA does not compose heavenly cantatas, nor even musical trash, although a lot of other trash has been produced through its help. If the ballyhoo had taken place a hundred years earlier, Johann Strauss would undoubtedly have composed a "Double Helix Polka." Our times would at best be capable of producing a dance that might perhaps be called the "dobblewobble."

   Even the early workers who experimented with living cells or tissues or with blood must have had a perception of the ghost-like presence of chemistry in all they were doing. They knew, of course, some chemistry and they knew that chemistry was the science of substances, of compounds. They must have realized that protoplasm was composed of compounds that, at least on one level of their existence, obeyed the laws of chemistry. But despite the early appearance of such men as Friedrich Miescher or Hoppe-Seyler I do not think that many bridges existed between biology and chemistry. The reasons why the early biologists kept their distance from the exact sciences do not all speak against the profundity of their perceptions. Reductionism had not yet entirely taken over their ways of thinking, as it has done now, and what one could call the technicalization of biology -- the wheels and the gears and the pulleys, the fuel, the lubricants, the templates, and so on -- had not yet won its shallow victory. There was still some reverence left, an awe before the everlasting mystery of life. Those that were good among these old men stepped softly.

    When I turn to the early stage of genetics I get the impression, perhaps mistakenly, that the initial exponents of this science were particularly unable or unwilling to think in terms of chemistry. Once the gene, as the unit of heredity, was defined and its localization in the chromosomes made probable, there was, of course, enough reason to assume that this unit was a substance or a conglomerate of substances and, thereby, subject to the scrutiny of the chemist. I am not able to determine how much speculation on the chemical nature of the gene did take place in the early days of genetics [see Bateson on these web-pages Click Here]; but I am sure that a geneticist, had he been pressed to reveal his thoughts, would have guessed that the gene may be a protein, for proteins were at that time the receptacles of all that was mysterious and refractory. Of course, the term "biological information" could not yet exist; those were the times of the log table, if not the abacus, and not of the computer.

     I am not sure if I am right in saying that if one wants to decipher an as yet unread language, simple, primitive texts may be more useful than complex, poetic documents. In the case of the chemical basis of heredity this is, however, unquestionably true. Without the use of phages and of microorganisms little could have been achieved in this regard. This stage was reached in the early forties, i.e., at the time when Oswald T. Avery and his collaborators began to work on the transformation of pneumococci, and Delbruck and Luria on the phages of E. coli, although Avery's laboratory was much more receptive to the application of chemistry than was the other group. It is really with Avery that there began what I have called, in my title, the chemical education of genetics. It is, perhaps, characteristic of the way in which science operates that the educator was neither a geneticist nor a chemist. The learning process was slow and weary: the card-carrying members of the guild and their assorted acolytes refused as long as possible, and even beyond that point, to take cognizance.

    The path of careful, conscientious, and responsible research that led Avery and collaborators (1) to the recognition that the hereditary units, the genes, were composed of DNA has been described excellently in the Dubos book (2). The amazing difficulties that this truly epochal observation experienced before being accepted have been narrated comprehensively in Olby's book (3). I should also like to refer to my reviews of the books of Dubos (4) and Olby (5).

    How profound the impression was that Avery's discovery made on me I have attempted to relate more than once (6,7). I shall return to it presently. One should have thought that if I, a simple chemist only distantly interested in the mechanisms of heredity, was so deeply moved by the sudden appearance of a giant bridge between chemistry and genetics, the practitioners of the latter science would have been alerted even more forcefully. I had, however, at that time the impression that this was far from being the case. In preparing the present essay I wanted to confirm the accuracy of my recollection. I went, therefore, back to the library and looked through a few genetics texts that were current in the period following the 1944 paper by the Avery group.

    The oldest book I consulted was the fourth edition, published in 1950, of a widely read introduction by Sinnot, Dunn, and Dobzhansky (8). The last two of the three authors, both Columbia professors, I had known very well during their lifetimes. The names Beadle, Delbruck, Lederberg, Luria, and Tatum appear in the index, and so does, more modestly, DNA as a component of chromosomes and some viruses. The name of Avery, however, does not appear; and so far as I can see -- I did not again read the entire 500 pages -- his discovery is not mentioned.

    In a textbook published three years later (9) Avery is listed in the index. My joyful exclamation was, however, stifled when I discovered that it was, alas, the wrong Avery. The "tetranucleotide theory of Levene" is discussed at length in ancient terms; but, again, no ripple of the wave of the future.

    The next candidate is a small monograph on the biochemistry of genetics, published in 1954 (10). As was to be expected of so intelligent an author as J. B. S. Haldane was, the discovery of pneumococcal transformation by DNA is mentioned (p. 49); but there is little evidence of an awareness of what Avery's discovery meant in terms of the chemical structure of DNA and its role in the chromosomes. The nucleoproteins of the nucleus are regarded as catalysts (p. 117), perhaps in fixing ATP (p. 126). If the first two books rate an F, this one would rate C+.

   My last witness is a book published in 1958 (11). Most of the standard names can be found in the index, but neither Avery nor Crick and Watson. One should have thought that enough time had then elapsed to digest the digestible, not to speak of the precooked, such as the double helix. Mention is made of transformation being induced by a "nucleic acid, of a specific type"; but that is about all. Fourteen years after his discovery, and three years after his death, Avery did not even rate honorable mention. One gets the impression that the tenaciously engrained conception of the classical gene acted as a vanishing cap for its real unraveler.

    The title I have chosen is, therefore, possibly wrong, and it should read: "How Genetics Refused to Get a Chemical Education." But "genetics" is, of course, as vague an entity as "the people"; and the collective is made up of all sorts of individuals, each one doing his own thing. Besides, it is a fortunate fact that amateurs often are better in advancing science than are the professionals. Nothing more deadening than being a specialist, an expert. You lecture before a perpetually somnolent audience -- the people change, but they are equally bored or obtuse -- or, if you are lucky, you teach in a workshop on a beautiful island, and you teach them to become as you are; whereas what a scientist ought to do is to teach others to become as different from himself as possible. Vive la difference! should be the battle cry. Instead, it is "like begets alike," until at the end dismal sociobiology takes over to tell us that you must be programmed in your genes to attend Asilomar [a place of scientific meetings]. Scientific life nowadays would be funny if it were not sad.

    So let me think of better times. I have been trying to recollect when I first heard of nucleic acids: probably during my University time, but I cannot have learned more than what I learned about insect pigments or anthocyanins. As a post-doc at Yale I saw, however, T. B. Johnson every day, and there the purines and pyrimidines made themselves known plentifully. As a young Assistent at the Bacteriology Department of the University of Berlin I earned a little extra money, writing abstracts for the Centralblatt, and one day I got the newly published book by P. A. Levene (12) for review. This was late in 1931 or at the beginning of 1932. I read it dutifully, but I do not remember any more what I said about it. The book certainly was not particularly pleasant to read, although I have kept my review copy to this day.

    In my own work I encountered DNA when Seymour Cohen studied the composition of rickettsiae (13) and together with him and Aaron Bendich we came across RNA in our work on the thromboplastic protein (14). What a job it was to do a spectrum on the Hilger spectrograph! Besides, the wet blanket of the tetranucleotide hypothesis extinguished all enthusiasm for these unpleasant laboratory curiosities. But at about the time when I wrote those two papers, deoxyribonucleic acid captivated my attention in a much more compelling manner.

   Was it in the at-that-time still pleasant dining room? Was it in one of the cheerless corridors of the College of Physicians and Surgeons? Anyway, somebody came and told me to read a paper by Avery in the Journal of Experimental Medicine. This was the article I mentioned before (1). Associations of thought normally cannot be reconstructed after the event, for they have the logic of dreams; but it was obvious to me that I must work on the chemistry of the nucleic acids. The road to take was, in fact, clearly delineated before my eyes; what I did not know was how to get to the beginning of the road. I knew that we had to find methods for the complete and accurate analysis of the nitrogenous components and the sugars of several DNA specimens widely separated as to their origins. Since most specimens would be difficult to come by, the methods, moreover, had to be applicable to minute amounts.

    The immediate problems, then, were (1) to develop procedures for the quantitative analysis of each of the purines and the pyrimidines present in a DNA; (2) to establish satisfactory balances in terms of total N and P; (3) to identify the sugar, or the sugars, present in a given nucleic acid; (4) to secure a variety of intact DNA specimens.

   Memory, unless it is committed to writing (and even then), is the most evanescent of gifts. How many are there still left who remember what it meant to determine the composition, let alone the exact composition, of a nucleic acid in, say, the year 1945? If you consider that at that time the quantitative analysis even of a protein could not be achieved -- and proteins had been studied much longer and more intensively than the nucleic acids --you will conclude, and rightly so, that nothing could be done for the nucleic acids. Levene's book epitomized the situation. For instance, on page 113 of this treatise the following statement will be found: "No methods exist for the quantitative determination of the individual purines when present in mixtures." For the qualitative isolation of the constituent purines, procedures requiring 50 g of nucleic acid are described, and the same is true of the pyrimidines.

    It is, therefore, perhaps not surprising that in the absence of any means of ascertaining the truth about the composition of the nucleic acids, a form of mock democracy was observed by the investigator who proclaimed: "All nucleic acid bases are equal." That some could be more equal than others did not even dawn on the archreductionist. This led to the baseless tetranucleotide theory {*Footnote: Much later some people came and told me that they never had believed in the tetranucleotide structure. This may be so; but by sitting solidly on one's haunches, while having hunches, one does not advance science.} I only regret that P. A. Levene did not call it the "Central Tetranucleotide Dogma," as sillier times would undoubtedly have done. To topple a dogma is more fun than to disprove a theory, for topple or disprove we certainly did.

    When in 1946, together with Ernst Vischer and Charlotte Green, I set out to develop a quantitative micromethod for the complete analysis first of DNA, and a little later also of RNA, we were favored by an unusual conjunction of lucky circumstances.

• First of all, the introduction of paper chromatography two years earlier, for the qualitative separation of amino acids (15) had shown the possibility of separating, and in most cases identifying unambiguously, minute quantities of organic substances.

• Secondly, at the end of the war, in the beginning of 1946, excellent photoelectric quartz spectrophotometers became available commercially, and we were among the early acquirers of a Beckman instrument. And

• thirdly, and most importantly, the possession by the purines and pyrimidines of specific and characteristic absorption spectra in the ultraviolet,

together with the arrangements mentioned just now, rendered feasible for the first time the development of an exact procedure for quantitative microestimation by paper chromatography. Our first preliminary note was submitted in April 1947 (16) and a detailed paper in June 1948 (17).

    One month later, we sent in two papers on the complete quantitative analysis of several DNA preparations: one dealt with the purines and pyrimidines of the DNA of calf thymus and beef spleen (18), the other with those of the DNA of tubercle bacilli and of yeast (19).

   One curious circumstance attending the publication of these papers deserves mention because it illustrates the ignorance about nucleic acids that then prevailed among the scientific elite. I had, at that time, already published something like 75 articles in the Journal of Biological Chemistry without ever having one sent back by the editor for clarification or revision. The papers about DNA composition, however, were returned to me with a particularly silly objection. How could I, the editor asked, express the composition of a DNA as moles of adenine or guanine, cytosine or thymine, per gram-atom of phosphorus, since the purines and pyrimidines did not contain any phosphorus? After I had repeated, in my answer to the editor, part of the introductory lecture on the nucleic acids, which at that time I was already giving to the first-year medical students at Columbia, we achieved grudging reconciliation.

    The emphasis that I placed on molar relationships underlines the fact that I approached the problem as a chemist. The phosphorus and nitrogen contents of a nucleic acid preparation can be ascertained by elementary analysis, and we did not rest satisfied until our analytical methods permitted recoveries of the total bases in the range of 96% to 98%. This is no longer done in experimental studies on DNA. Sometimes complete nucleotide sequences are now published without any proof that these correspond to the total P, N, and purine and pyrimidine contents of the specimen. In this way, half-truths are piled on quarter-truths until one day the entire structures will collapse. In my opinion, molecular biology is disregarding chemistry at its own peril.

    Another utopian attempt of ours was the investigation of the nature of the deoxy sugar in every DNA specimen prepared by us. It is true, nothing but deoxyribose was found; but there was no reason to assume beforehand this to be the case. In other words, someone had to do what modern scientists would consider as "dirty work," and we were not loath to do it.

     I have recently described the path that led us to our present view of the chemical nature of DNA (7), and I should not wish to repeat myself here. When the time had come for me to attempt a first summary (20;[1950]), this is what I wrote:

     I am not a historian of science -- if there is such a thing -- and I am, therefore, not sure that I am correct in saying that this is among the early statements concerning chemically encoded biological information. My claim is perhaps strengthened by another passage from the same review, affirming the biological importance of nucleotide sequence:

One other short paragraph of the same article (20) carried the seeds of the future. It reads:

How this statement came to be inserted into the galley, proofs of the review article (20), I have recounted in my forthcoming book (7). Our first results, marred by the initial necessity of determining the purines and pyrimidines separately and by an indirect procedure of demonstrating the separated spots on the filter paper, did not lead obviously to such a conclusion. Without the complete balance of recoveries in terms of nucleotide phosphorus, which we established in all our analyses, we should never have come to the recognition of the remarkable pairing rules. The first two observations we made on the basis of these balances were (1) the recoveries of purines were invariably much higher than those of pyrimidines: a difference attributable to the different hydrolysis procedures then employed for the liberation of these two classes of cornpounds; (2) even in the first studies on DNA of beef tissue (18), and despite the higher yield of purities that I have mentioned, the molar ratios of adenine to guanine were very similar to those of thymine to cytosine: the average for A/G was 1. 3 and that for T/C 1.4.

    When I gave a series of lectures in 1949, I mentioned these and related observations, but when it came to rewriting them in the form of the Experientia review (20), I hesitated first to emphasize any compositional regularities, owing perhaps to my skeptical and antidogmatic character. But in the meantime we had begun to improve our initial methods considerably by introducing formic acid hydrolysis for the simultaneous liberation of all nitrogenous constituents and by using a suitable UV lamp for the demonstration of the separated adsorption zones on the filter strip. The rapidly accumulating new results encouraged me to insert the well-known paragraph.

     Is the title that I have chosen for this brief account justified? Did genetics, did biology in general, receive a chemical education during the period through which I have lived? To what extent did my own laboratory participate in this effort? The answer to the last question I shall have to leave to others. But as concerns chemistry supplying the foundation of the life sciences in our days, not just an underpinning, the answer is Yes. Even the most recalcitrant geneticist, even the most nebulous of immunologists, can no longer disregard the very science of substances that is chemistry. The victory may have been, however, a Pyrrhic one. In teaching them the nomenclature, we may not have taught them the skill. Compounds that had to be prepared in the laboratory, in a painful struggle for purity, now are supplied at great cost by sloppy merchants, painlessly, dirtily. For this reason, large areas of molecular biology appear to operate in a cloud of ambiguity that may turn out to be lethal.

    In any event, it looks as if the smattering of chemistry acquired by biology has killed biochemistry. Does anyone care? Has the history of the decline and fall of a science ever been written? Can it be, except at a distance of many centuries?

     Daily we are told about the great success of one science or the advances made by another. What is success in science, what is advance? Is there a final goal, is there such a thing as a better truth? Like everything else, the sciences are governed by the principle of change; but I should hesitate to replace the letter G in that word by the letter C, as was done by the late Jacques Monod. In fact, Change and Inertia would be a better title than Chance and Necessity [the title of a book by Monod]. The waves roll on and break, and then there come other waves. Each is different, although they all are waves. We try to adapt ourselves to whatever force carries us at any given time. Confident of leaving our imprint on them, we are actually shaped by the wave of the moment.

REFERENCES

1. AVERY, 0. T., C. M. MACLEOD & M. McCARTY. 1944. Studies on the chemical nature of the substance inducing transformation of pneumococcal types. Induction of transformation by a desoxyribonucleic acid fraction isolated from pneumococcus Type 111. J. Exp. Med. 79: 137-158.

2. DUBOS, R. J. 1976. The Professor, the Institute, and DNA. Rockefeller University Press. New York, N.Y.

3. OLBY, R. 1974. The Path to the Double Helix. University of Washington Press. Seattle.

4. CHARGAFF, E. 1977. Experimenta lucifera. Nature 266: 780-781.

5. CHARGAFF, E. 1976. Review of Olby (Reference 3). Perspect. Biol. Med. 19: 289-290.

6. CHARGAFF, E. 197l. Preface to a Grammar of Biology. Science 177: 637-642.

7. CHARGAFF, E. 1978. Heraclitean Fire. Rockefeller University Press. New York, NY.

8. SINNOTT, E. W., L. C. DUNN & T. DOBZHANSKY. 1950. Principles of Genetics. 4th edit. McGraw-Hill Book Company. New York, N.Y.

9. HOVANITZ, W. 1953. Textbook of Genetics. Elsevier. New York, N.Y.

10. HALDANE, J. B. S. 1954. The Biochemistry of Genetics. Allen & Unwin. London.

11. SRB, A. M. & R. D. OWEN. 1958. General Genetics. Freeman. San Francisco.

12. LEVENE, P. A., & L. W. BASS. 1931. Nucleic Acids. Chemical Catalog Co., NewYork, N.Y.

13. COHEN, S. S. & E. CHARGAFF. 1944. Studies on the composition of rickettsia prowazeki. J. Biol. Chem. 154: 691-704.

14. CHARGAFF, E., A. BENDICH & S. S. COHEN. 1944. The thromboplastic protein: Structure, properties, disintegration. J. Biol. Chem. I56: 161-178.

15. CONSDEN, R., A. H. GORDON & A. J. P. MARTIN. 1944. Qualitative analysis of proteins: A partition chromatographic method using paper. Biochem. J. 38: 224-232.

16. VISCHER, E. & E. CHARGAFF. 1947. The separation and characterization of purines in minute amounts of nucleic acid hydrolysates. J. Biol. Chem. 168: 781-782.

17. VISCHER, E. & E. CHARGAFF. 1948. The separation and quantitative estimation of purines and pyrimidines in minute amounts. J. Biol. Chem. 176: 703-714.

18. CHARGAFF, E., E. VISCHER, R. DONIGER, C. GREEN & F. MISANI. 1949. The composition of the desoxypentosc nucleic acids of thymus and spleen. J. Biol. Chem. 177: 405-416.

19. VISCHER, E., S. ZAMENHOF & E. CHARGAFF. 1949. Microbial nucleic acids: The desoxypentose nucleic acids of avian tubercle bacilli and yeast. J. Biol. Chem.177: 429-438.

20. CHARGAFF, E. 1950. Chemical specificity of nucleic acids and mechanism of their enzymatic degradation. Experientia 6: 201-209.

21. ELSON, D. & E. CHARGAFF. 1955. Evidence of common regularities in the composition of pentose nucleic acids. Biochim. Biophys. Acta 17: 367-376.

22. KARKAS, J. D., R. RUDNER & E. CHARGAFF. 1968. Separation of B. subtilis DNA into complementary strands, II. Template functions and composition as determined by transcription with RNA polymcrase. Proc. Nat. Acad. Sci. U.S.A. 60: 915-920.

2 3. RUDNER, R., J. D. KARKAS & E. CHARGAFF. 1968. Separation of B. subtilis DNA into complementary strands, 111. Direct analysis. Proc. Nat. Acad. Sci. U.S.A. 60: 921-922.

24. KARKAS, J. D., R. RUDNER & E. CHARGAFF. 1970. Template properties of complementary fractions of denatured microbial deoxyribonucleic acids. Proc. Nat. Acad. Sci. U.S.A. 65: 1049-1056.

DISCUSSION OF THE PAPER [after Chargaff's address].