This page contains the full text of the following article:
Athel Cornish-Bowden (1998)
Two Centuries of Catalysis
Journal of Biosciences
With the centenary of the discovery of cell-free fermentation (Buchner, 1897) we are entering the second century of scientific biochemistry, based on the idea that life processes are no more than chemistry, uncomplicated by notions of vitalism. At about the same time we are also entering the third century of the study of catalysis — a surprising statement, perhaps, as catalysis is usually thought to have started with Jöns Jacob Berzelius in 1836.
More than 40 years before Berzelius, however, Elizabeth
Fulhame (1794) published a book entitled An Essay on
Combustion in which she propounded ideas that are
recognizable today as the first suggestions of catalysis.
Her starting point was the hope of finding a satisfactory
way of staining cloth with heavy metals under the influence
The possibility of making cloths of gold, silver, and other metals, by chymical processes, occurred to me in the year 1780.
Not for the first time, the ideas of an amateur, and a woman at that, were not immediately taken very seriously by her husband and his friends:
The project being mentioned to Doctor Fulhame, and some friends, was deemed improbable.
Nevertheless, she was right:
However, after some time, I had the satisfaction of realizing the idea, in some degree, by experiment.
Mrs Fulhame’s claim to be regarded as the first to describe catalysis rests on the principal conclusion that she drew from her experiments, that
the hydrogen of water is the only substance, that restores oxygenated bodies to their combustible state; and that water is the only source of the oxygen, which oxygenates combustible bodies.
She believed, in other words, that water acts in oxidation-reduction reactions both as a reducing and as an oxidizing agent, being restored to its original state at the end of the process. This is rather an extreme view, of course, as oxidation-reduction reactions that do not involve water certainly exist. Later on Berzelius adopted a position at the opposite extreme, when he described catalysis as an
only rarely observed force which is probably active in the formation of organic substances.
Little is known about Mrs Fulhame’s life beyond the fact that she appears to have been the wife of Thomas Fulhame, a London doctor. Of her work we know only of her book and the fact that she was in contact with some of the well known scientists of her time, such as Joseph Priestley. Her book appeared in both British and American editions; it is virtually unobtainable today*, but a detailed synopsis of it in French was prepared by Coindet (1798), and this is certainly easier to locate than the book itself.
*This was true when written, but is no longer true today. The book has been reprinted and is available from
Amazon and elsewhere for a very reasonable price. Although the publisher warns that
Due to the very old age and scarcity of this book, many of the pages may be hard to read due to the blurring of the original text, possible missing pages, missing
text and other issues beyond our control, this is excessively pessimistic: certainly some pages are not very clean, but all are readable and I haven't detected any missing text. (Note added 14 May 2010.)
A little more detail about Elizabeth Fulhame may be found in a recent chapter (Laidler and Cornish-Bowden, 1997) in a book to commemorate the centenary of Eduard Buchner’s discovery of cell-free fermentation (Cornish-Bowden, 1997). This includes a new translation of Buchner’s classic paper into English (Buchner, 1997) and sets out to relate it to the present state of knowledge of the kinetics of multi-enzyme systems.
As Herbert Friedmann (1997) discusses in a masterly chapter in the same book, the century that separated Mrs Fulhame from Buchner was dominated by arguments over vitalism. The story is far from being a simple one, with clearly identifiable villains being overcome by clearly identifiable heroes. Some of the most vigorous defence of the vitalist position came from excellent scientists, such as Berzelius and Louis Pasteur, and some of the fiercest attacks on the vitalists, such as those of Justus von Liebig and Friedrich Wöhler, were misconceived. Moreover, one cannot see it as a simple development from an incorrect vitalistic view of physiology to a correct one based on chemistry.
Much of the controversy was focussed on alcoholic fermentation, especially in yeast, and one can identify two extreme positions, equally wrong-headed from our present point of view. Vitalists such as Pasteur argued that yeast was a living organism and that the chemical transformations that it brought about could not be dissociated from its living nature. However, we should be wrong to criticize this view too strongly. First of all, it was soundly based on experimental evidence: Pasteur made careful and strenuous efforts to observe fermentation in cell-free extracts of yeast, but he did not succeed. (Arthur Harden, one of the giants of fermentation research, showed afterwards that Pasteur’s failure was not due to any technical inadequacy of his procedure but to the fact that the Paris yeast that he used was not suitable for preparing extracts.) Perhaps more important, Pasteur’s view ought not to be contrasted with the modern view that yeast is a living organism in which chemical reactions take place, but with an anti-vitalist view even more extreme than vitalism.
The leading opponents of vitalism in the mid-19th Century, such as Liebig and Wöhler, were not content just to argue that alcoholic fermentation was a chemical process; they rejected the whole idea that yeast could be a living organism. They abused the influential journal that they controlled, Annalen der Pharmacie (later Liebig’s Annalen), in a way that we should not regard as acceptable today, ridiculing scientists who claimed to have observed growing yeast cells under the microscope. Wöhler’s discovery that urea could be synthesized by heating the inorganic compound ammonium cyanate had led him and his associates to an excessively simple view, seeing fermentation as a simple chemical process with no need of the organization that a living organism could provide.
Friedmann’s (1997) account of all this makes fascinating reading, but it is difficult to escape the conclusion that the development of understanding of fermentation in the 19th Century was a very muddled business that resists a simple presentation, in part, as I have indicated, because the distinction between heroes and villains is impossible to make. Great scientists who are remembered for other reasons proved to be on the wrong side of the vitalism debate for at least part of the time, or they were on the right side for the wrong reasons.
It is far easier to see the 100 years that followed Buchner in straightforward Whiggish terms as continuous progress towards the triumph of truth and reason over error. This is, of course, a grotesque oversimplification, ignoring the many wrong trails followed during this century, but these are easy to ignore, because the people who espoused them are not remembered for any positive contributions that they made. We can forget the cyclol theory of protein structure, for example, because we do not need to remember its proponents for any other reason.
Richard Willstätter is perhaps the most obvious exception. Resolutely opposed to any idea that enzymes might be proteins, he also refused to believe that cell-free fermentation mimicked the process in the cell. He was highly influential in his time, but he is now remembered more by chemists than by biochemists: authors of modern biochemistry textbooks do not remember him for the work on chlorophyll that led to the Nobel Prize for Chemistry in 1915; if they remember him at all it is for having cast scorn on the first reports of crystalline proteins. We now know, of course, that not all biological catalysis is mediated by proteins, but the existence of catalytic RNA hardly alters the fact that the great majority of enzymes are proteins.
If we accept that the discovery of cell-free fermentation was the turning point in the history of biochemistry, then we need to ask why it was so important. Sounding the death knell of vitalism was crucial, of course, and it opened up whole areas of study that could not have been usefully studied before. Buchner thought that his zymase was a single enzyme, the catalyst for the whole fermentation pathway, but this oversimplification was of little importance: once it was clear that a catalyst existed that acted in accordance with chemical laws, it became reasonable to try to separate it into its components, and this started soon afterwards. With the component enzymes came the component metabolites, so that fermentation ceased to be a black box in which glucose was converted into ethanol and carbon dioxide, but was seen to be a pathway proceeding in steps. Thus metabolism became, along with enzymology, one of the dominant themes of 20th Century biochemistry.
Although enzymology existed a little earlier, its development was greatly stimulated by Buchner’s discovery, and the first decades of the 20th Century saw the groundwork laid for the study of the kinetics and mechanisms of enzyme action, most notably by Leonor Michaelis and his collaborators. As a digression, it is interesting to notice Michaelis’s output and to realize that very high publication rates are not as recent a phenomenon as we may think. His most famous paper (Michaelis and Menten, 1913) was one of ten he published in 1913. Rather a lean year by his standards, this was the least productive of the five years from 1910 to 1914, which saw 94 publications, including five books. By no means all of the other 93 have been forgotten, either. Of his 23 publications in 1914, for example, two papers (Michaelis and Pechstein, 1914; Michaelis and Rona, 1914) established the basic ideas of enzyme inhibition, and his book Die Wasserstoffionenkonzentration (Michaelis, 1914) became the standard treatment of pH, buffers, etc.; all of these are still cited occasionally. Nor was Michaelis alone in his high publication rate: Victor Henri, who anticipated him in some respects in our understanding of enzyme action (Henri, 1901), was the author of more than 500 papers, and according to Boyde (1980) his other contributions were always in the first rank.
Before leaving Michaelis it is perhaps appropriate to mention some of his collaborators in the early years of the 20th Century. Three names occur frequently as co-authors: Peter Rona, Heinrich Davidsohn and H. Pechstein. Of the last two I know nothing apart from their published work, but Peter Rona was a chemist in the hospital where Michaelis worked in Berlin. He was his friend as well as his collaborator, and together they set up the laboratory in which the research was done.
Michaelis’s most famous
collaborator, Maud Menten, appeared on only one paper,
where her name appears rather enigmatically as
Maud L. Menten, and she seems to have spent only a short
time in Berlin. Despite her German-looking surname she was
Canadian, one of the first Canadian women to receive a
doctorate in medicine, and she worked mainly as a
pathologist in the USA. The L. stands for Leonora, and my
computer program for fitting enzyme kinetic data
(Cornish-Bowden, 1995) is named Leonora to commemorate both
her and Leonor Michaelis. As far as I have been able to
discover the similarity of their Christian names is no more
than a coincidence.
The first three-quarters of the century after Buchner can be seen as the period of consolidation in which the edifice that we now recognize as the science of biochemistry was constructed. More recently the main action has moved into the genetic interface between biochemistry and the other life sciences and the borderlines between these sciences are becoming blurred. Within a few years we expect to know the entire sequence of the human genome, and if we believe some commentators we may think that everything worth knowing about the classical foundations of biochemistry, metabolism and enzymology, is now known, and that only people unable to keep up with the changes of fashion are still interested in these.
In reality, however, a great deal of life remains in metabolism and enzymology, and it is only recently that there has been much attempt to unite them into a single subject. Everyone knows that the great majority of intracellular enzymes act as components of metabolic pathways, but very few of the enormous number of papers on enzymes have paid any attention to this central fact of their existence; on the contrary, most enzymes have been studied with a view to understanding their mechanisms of action, and with this in mind the first thing that is normally done when studying an enzyme is to purify it, or in other words to take it out of its physiological context. This is necessary, of course, as a first step, as it would have been difficult to have learned much about enzymes if they had never been purified, but it is also important to recognize that it is only a first step and that the great bulk of research on the physiological role of enzymes lies in the future.
The first serious attempts to understand how enzymes operate as components of metabolic pathways were probably the computer modelling studies of Garfinkel and Hess (1964). Given the feeble computing resources they had at their disposal — they needed 30 minutes of computer time to simulate 75 ms of real time, and they had to overcome such surprising difficulties (to our eyes) as having to run their programs in machines that could not easily represent quantities greater than 1 — the progress that they made was impressive. Ultimately, however, computer modelling is a sort of sophisticated experimentation, and like any experimentation it needs a theoretical background if its results are to make sense. There were various attempts to construct such a background, of which the most successful has been the metabolic control analysis that grew out of the work of Kacser and Burns (1973). This is not the place to go into a detailed description of this approach, but a good modern one is given by Fell (1996); suffice it to say that we now have a language appropriate for describing and analysing the kinetic properties of physiological systems — pathways, cells, organs, whole organisms and even ecosystems — so that we can begin to define metabolic regulation in terms of the properties of the whole regulated system. We may hope that within the next decade we may have a modern comprehensive theory of metabolic regulation that puts the single-enzyme mechanisms — allosteric regulation, cooperativity, etc — of a quarter of a century ago into a multi-enzyme context, and some steps in this direction have been taken already (Hofmeyr and Cornish-Bowden 1991).
Why do we need such a theory? If we read about drug design in the modern biotechnology literature, such as the recent special supplement of Nature (1996), we find a great deal of discussion of combinatorial chemistry, bioinformatics, methods of drug screening and other fashionable topics, but barely a mention of metabolism — no recognition, in other words, that drugs act by modifying metabolism. At the same time the success rate obtained with these modern methods is not greatly different from the rates that were common thirty years ago, with one commercially valuable product for 5000 compounds tested. This is not very impressive, and it is hard to escape the conclusion that the disappointing performance of this and other areas of biotechnology is related to the lack of attention paid to metabolism and enzymology. What drugs do is to modify metabolism, and they do it by modifying enzyme properties. It seems self-evident, therefore, that before one can hope to predict how a pharmacological agent will act in the living organism one needs to understand the metabolic behaviour of the organism in the absence of the agent and the dependence of this behaviour on the activities of the relevant enzymes. It is this type of understanding that we can expect to be supplied by an adequate theory of metabolic regulation.
Until it is tried in a serious way we cannot be certain, of course, that drug design based on the principles of metabolic regulation will be any more successful than current practice, but it is reasonable to hope that it will be, and some results from computer modelling, sophisticated experimentation as I described it above, suggest that this is so. Efforts to find drugs to control or cure African sleeping sickness, the disease transmitted by Trypanosoma brucei, have mainly focussed on the transport of glucose into the parasite from the host blood, and we have used a combination of metabolic control analysis and computer modelling (Eisenthal and Cornish-Bowden, 1998) to study how well this is likely to work. T. brucei has an extremely high glycolytic rate, and to a first approximation one can draw a metabolic scheme consisting of 20 processes (mainly enzyme-catalysed reactions, with a few transport steps) that account for most of its metabolic activity. One might interpret this as indicating the existence of 20 potential targets for enzyme inhibitors to act as drugs, but before discussing this it will be helpful to indicate why T. brucei is the metabolic modeller’s dream organism.
Metabolic modelling usually involves a considerable amount of guesswork, because it usually turns out that only a minority of the relevant enzymes (if any) have been studied in the relevant organism, and when they have been studied it is usually necessary to combine data from different groups made under arbitrarily different conditions, and it is rare for any data to be available about the kinetics of the reverse reactions or even about product inhibition. As a result of an extended programme of research carried out by Fred Opperdoes and his colleagues in Brussels (see Bakker et al., 1997), T. brucei is an exception to all of this: nearly all of its glycolytic enzymes have been studied kinetically by a single group under comparable conditions, and the data include information about the reverse reactions. One can thus address questions to the computer with a reasonable expectation that the answers will have some relevance to what happens in the living parasite. In one respect at least, this is confirmed, as the computer model proved able to predict with almost best-fit precision observations for the transition from anaerobic to aerobic conditions that had not been used during its construction.
Let us now return to the 20 potential drug targets in T. brucei. When studied in the computer it appeared that only two of these, the glucose and pyruvate transporters, had much likelihood of being useful targets. If the aim is to kill the parasite by decreasing the glycolytic flux to a level insufficient for life then inhibiting glucose transport is the only hope, but it turns out to be a rather forlorn hope because very high levels of inhibition (with the risks of deleterious side effects) would be needed to achieve appreciable depression of metabolism. The other candidate, the pyruvate transporter, proves to be much more promising. Inhibiting it uncompetitively should have almost no detectable effect on the glycolytic flux, but a huge effect on the internal pyruvate concentration, and it is not unreasonable to hope that the organism would be unable to cope with this. (Why uncompetitively, incidentally, given that most people think only of competitive inhibition? Because in general competitive inhibition has little effects of any kind in a living organism at any reasonable ratio of inhibitor concentration to inhibition constant. An essential point that is obvious but is nearly always forgotten in efforts to design drugs is that anything that can compete with the substrate for the active site of an enzyme is something that the substrate can compete with.)
There are other examples that could be used similarly to illustrate why more attention will need to be paid in the future to kinetics and metabolism if the promises of biotechnology are to be realized. The point is that no amount of combinatorial chemistry, gene sequencing, bioinformatics etc. could reveal that pyruvate transport is likely to be a weak point (possibly the only weak point) for attacking T. brucei, and these fashionable approaches are equally unlikely to lead to ways of modifying organisms to produce larger quantities of commercially desirable metabolites. To modify metabolism it is necessary to understand metabolism and kinetics, especially multi-enzyme kinetics.
There are many other important questions in the organization of metabolism that still need to be answered. For example, since the discovery of protein kinases and phosphatases it has become clear that many enzymes exist as pairs of proteins, one active and the other inactive, that are interconvertible by covalent modification reactions catalysed by other enzymes. The existence of such cycles is not hard to explain, as they permit a degree of sensitivity to stimuli far greater than can be provided by the classical regulatory mechanisms like allosteric inhibition and cooperativity. However, there are many cases in which the enzymes that catalyse the covalent modification reactions are themselves members of covalent modification cycles, and sometimes there are more than two nested cycles of this kind (Cárdenas, 1997). For the moment there has been no satisfactory analysis of the properties of such cycles within cycles that could explain their existence. Although some authors have suggested that the explanation lies in the high sensitivity to signals that they allow this is very difficult to believe, given the already enormous sensitivity possible with a single cycle. This is but one of many outstanding questions in metabolic regulation that we may hope that research in the third century of studies of catalysis will solve.