Dr Henrik Kacser (1918-1995)

This page contains an obituary of Henrik Kacser by David Fell (School of Biological and Molecular Sciences, Oxford Brookes University, Oxford OX3 0BP, England). It is reproduced, with minor modifications, from the obituary published in The Biochemist 17, No. 3, 26-27 (1995), and also in the Journal of Theoretical Biology 182, 193–194 (1996).

Was it because Henrik was not a biochemist that he came to undermine so comprehensively one of the barely challenged dogmas of metabolic biochemistry: that pathways are controlled by a rate-limiting step or pacemaker? Probably it made little difference for, whatever the topic, he took no argument on trust; he expected to be given convincing reasons, and as his colleagues and students knew, he had an unfailing ability to find the weak points in an argument or its presentation. In his own subject of biochemical genetics, the phenomenon of dominance of wild-type over mutant genes was generally regarded as requiring little explanation, yet his work has shown that it can be accounted for by the system properties of metabolism.

Henrik was born in Romania of Austro-Hungarian parents who later moved to Berlin where Henrik went to school. Before the Second World War, he went to Queen’s University, Belfast to study chemistry, and specialized in physical chemistry as a postgraduate student. In 1952 he moved to the University of Edinburgh as a Nuffield Fellow under a scheme to introduce physical scientists into biology. He took the Diploma of Animal Genetics course and was invited to stay on in the Department of Genetics, becoming a lecturer in 1955.

Although he taught genetics, his theoretical and experimental research was influenced by his background in physical chemistry. By the end of the 1960s, he and Jim Burns were working on the theory of how the rates of metabolic pathways were affected by changes in the amounts or activities of pathway enzymes. Their landmark paper, The control of flux (Kacser & Burns, 1973), was presented before a sceptical audience of experimental biochemists in Oxford in 1973. In it they show that the expectation that a metabolic pathway will be controlled by a single pacemaker reaction is a fallacy, and most of the experimental criteria used in the supposed identification of such steps are misleading. Instead, varying amounts of control can be distributed over the enzymes of the pathway, but this is a property of the metabolic system as a whole and cannot be predicted from the characteristics of the enzymes in isolation. As we prepared it for re-publication (Kacser et al., 1995) in conjunction with the colloquium held at the Sussex Meeting of the Biochemical Society in December 1994 to mark the 21 years since its first appearance, I was struck by the confidence and clarity of the argument. This is a theory that has been thoroughly worked through and presented by authors who are comfortable with the ideas. It exhibits Henrik’s characteristics as a scientist: the writing is clear and precise; the presentation is logical and supported by as much mathematics as is necessary, but the mathematics is always secondary to the physical and biological implications.

Recognition of the article was slow to come. Perhaps this was not surprising: it challenged the qualitative, reductionist basis of the universal approach to control in metabolic biochemistry, stated that a quantitative systems approach was necessary, and it contained equations. It is also true that Henrik did not follow the article up rapidly himself. Some of the experimental work he cited in the 1973 paper did not appear until the 1980s; some is still unpublished. Also slow to be revealed was the full meaning of the footnote in The control of flux: the implication of this for the problem of dominance and its evolution will be dealt with in a separate publication; this companion paper, The molecular basis of dominance (Kacser & Burns, 1981) appeared only eight years later. The connection was that if the flux–enzyme relationship is quasi-hyperbolic, and if, for most enzymes, the wild-type diploid level of enzyme activity occurs where the curve is levelling out, then a heterozygote of the wild-type with a null mutant will have half the enzyme activity but will not exhibit a noticeably reduced flux. Therefore the wild type appears dominant and the mutant recessive because of the system characteristics of a metabolic pathway, not because of special evolved properties of the types proposed by Fisher or Wright. This interpretation of dominance has been recently supported by results showing that it as common in artificial diploids of a normally strictly haploid organism as it is in habitual diploids (Orr, 1991). A fuller discussion of The molecular basis of dominance appears elsewhere in this volume (Porteous, 1996).

Biochemical interest in the ideas expressed in The control of flux started to grow in the 1980s, particularly with its experimental applications in Amsterdam to oxidative phosphorylation, urea synthesis and gluconeogenesis. At this time, because the theory of Kacser and Burns and the simultaneous but independent work carried out by Reinhart Heinrich and Tom Rapoport in Berlin were compatible, a common terminology and set of symbols was agreed for the new field of Metabolic Control Analysis.

Since then, the growth of theoretical and experimental studies of the ideas originally presented in The control of flux has continued to accelerate, though it is surprising how many of the problems and applications are anticipated in the original article. Henrik’s achievement has been recognized by his election to the Royal Society of Edinburgh in 1990, by an Honorary Doctorate of the University of Bordeaux II in 1993, and by the 21st anniversary colloquium mentioned above.

Henrik continued his research after his official retirement in 1988 and was particularly concerned with the problems that Metabolic Control Analysis does not appear to solve, such as how to predict the effects of large changes in enzyme amounts. Recent papers, in collaboration with Rankin Small and Luis Acerenza, have shown that the prospects for achieving large increases in flux by changing the activity of a single enzyme are poor (Small & Kacser, 1993ab), but that a coordinated set of changes, designed by their Universal Method, could make large changes without catastrophic perturbations of the rest of metabolism (Kacser & Acerenza, 1993). At the time of his death, Henrik still ran an active laboratory, had two large grants supporting his work and continued to produce original scientific ideas.

I thank Professors Geoffrey Beale and Graham Bulfield for allowing me to use material from their obituary notice and Henrik’s wife, Elaine Kacser, for the photograph.