This page is the third of a series containing the article
Putting the Systems back into Systems Biology by
Athel Cornish-Bowden, published in Perspectives in Biology and Medicine (2006) 49, 475–489.
A PDF file is also available.
If someone were to analyze current notions and fashionable catchwords, he would find
high on the list.
Bertalanffy, L. von (1969) General system theory, New York: George Braziller
2The quotation marks are to stress that much of the work that currently goes under
the name of
systems biology hardly merits this name.
3Bertalanffy’s capital letters.
After a very slow start, the ideas of Kacser and Burns (1973) are gradually coming to be accepted as the appropriate
basis for analyzing the kinetic behavior of multienzyme systems. What, however, does this acceptance of one kind of
systemic view of biology have in common with the sudden vogue for
systems biology on the one hand, or with the
older ideas of systems theory advocated by Ludwig von Bertalanffy and Robert Rosen on the other? At the moment it would
seem that there is remarkably little connection with either. Bertalanffy in particular laid great stress on the need to
see systems as wholes, and he was critical of what he saw of the ever-increasing specialization of modern science,
necessitated by the enormous amount of data, yet the very popularity of
systems biology is itself a
product of the amount of data generated by genome sequencing—enormous beyond anything that Bertalanffy encountered. His
major concern was to emphasize the essential Unity of Science3, which he thought was granted
not by a
reduction of all sciences to physics and chemistry, but by the structural uniformities of the different levels of
reality (Bertalanffy, 1969). He was very skeptical of reductionist ideas, and these played little part in his
thinking. In practice it has proved very difficult for others to build on any details of his work.
When I started thinking about this essay in 2003 I was already thinking that
systems biology was a vogue
term, but the vogue that existed then was as nothing compared to what it has become. More than half of the 600 or so
publications one can find by searching for this combination of words at the end of 2005 date from 2005, and much more
than half of the remainder from 2004. This sudden apparent enthusiasm for systemic ideas after many years in the
wilderness must, however, offer little comfort to those who have thought in terms of systems for a long time, because
systems biology in current practice is not easy to distinguish from old-style reductionist biochemistry applied
on an ever-larger scale. The main novelty is the recognition that many aspects of cell biology have to be understood in
terms of interactions between two or more proteins, or two or more other entities: this is certainly an advance on the
way of thinking that characterized much of classical biochemistry, but it is still far from an appreciation of systems
Cornish-Bowden, A. and M. L. Cárdenas (2001b) Information transfer in metabolic pathways: effects of irreversible steps in computer models, Eur. J. Biochem. 268, 6616–6624
A simple example of the importance of system-level thinking is provided by feedback inhibition of a biosynthetic
pathway by its end-product, a type of metabolic regulation described in any textbook of biochemistry. For example,
aspartokinase (which in some organisms is a mixture of two or more isoenzymes with different regulatory properties)
catalyzes the first step in the interconversion of the aminoacids aspartate and lysine, and it is inhibited by lysine.
As drawn in most textbooks, the pathway ends there, at lysine, but that representation leaves much unclear, because it
fails to indicate why the cell needs to make lysine in the first place, or what purpose is served by having the first
enzyme in the pathway inhibited by lysine. The point that is left unstated, and that some will consider too obvious to
need stating, is that lysine is needed for protein synthesis; more generally, the
end-product in almost any
biosynthetic pathway is not the end of anything, but just the link between two processes. People who claim that that is
obvious may then go on to claim that it is also obvious that feedback inhibition of aspartokinase is needed to ensure
that lysine is synthesized at the rate needed for protein synthesis, neither faster nor slower. This second point is
not only not obvious; it is not even true, because if regulation of rates were the only thing needed it could be
perfectly well satisfied by the ordinary product inhibition backwards through the pathway that arises from the everyday
properties of virtually all enzymes. As long as rates are the only concern no feedback loop is needed, but it becomes
absolutely necessary when one recognizes that metabolite concentrations need to be taken into account as well: without
feedback inhibition (and other classical regulatory mechanisms) flux control would be achieved at the expense of huge
uncontrolled variations in the concentrations of intermediates (Cornish-Bowden and Cárdenas, 2001b); these would
normally be very harmful, even lethal, to the organism, and explain why suppressing regulatory mechanisms is not
normally a useful way of achieving biotechnological aims. All of this can be summarized by saying that the regulation
of most biosynthetic metabolic pathways can be understood in terms of ideas of supply and demand, with the benefit that
in general economic laws work much better in metabolism than they do in economics (Hofmeyr and Cornish-Bowden, 2000).
Savageau, M. A. (1990) Biochemical systems theory: alternative views of metabolic control. In Control of metabolic processes (ed. A. Cornish-Bowden and M. L. Cárdenas), 6987. New York: Plenum
Woese, C. R. (2004) A new biology for a new century, Microb. Molec. Biol. Rev. 68, 173–186
What about the influence of Bertalanffy’s view of systems on Kacser and Burns and their more recent followers? As
long as one connes oneself to general remarks there may seem to be some influence, but when one examines the major
modern systemic approaches in detail they prove to be entirely mechanistic. Carl Woese (2004) makes a useful
distinction between what he calls empirical reductionism and fundamentalist reductionism, the former being purely
methodological, a way of analyzing systems into their component parts in order to understand them better. As we have
seen, reductionist ideas played little part in Bertalanffy’s thinking, but Kacser and Burns, as well as Heinrich
Rapoport and Savageau, are empirical reductionists in Woese’s sense, as they explain the properties of systems
essentially as the sums of the properties of their components. Savageau (1990), however, makes the valuable point that
any respectable reductionist is also a reconstructionist: it is not enough to show how a system can be reduced
to its components; one should also show that the components can be put together again to make a functional whole. He
adds, however, that
the problem is that the reconstructionist phase of this program is seldom carried out.
By contrast, fundamentalist reductionism is essentially metaphysical, a belief about the nature of living organisms, that they can be completely understood in terms of the properties of their components. Anything else may be, and has been, derided as a return to vitalism, but Rosen in particular has argued very strongly that there is a third way of interpreting life that is neither vitalism nor exclusively mechanism, but is complexity. Before discussing Rosen’s view, however, a detour is necessary to outline Erwin Schrödinger’s question about the nature of life, and his efforts to answer it.