The Screening Hypothesis - a new explanation of secondary product diversity and function

The hypothesis

The hypothesis is composed of two parts.
1. The identification of a fundamental constraint that must be a very significant factor in the evolution of secondary metabolism.
Potent, specific biological activity is a rare property for a molecule to possess - that is why large screening programmes are needed find useful biological activity. The low frequency of potent, specific biological activity is a consequence of the specificity of ligand/binding site interactions. Some organisms can increased their fitness by making and exploiting biological active molecules however the low frequency of evolving new molecules places severe constraints on the evolution of the biochemical pathways leading to such products.
 2. A proposal as to how secondary metabolism might have evolved with such a constraint - including a prediction as to the metabolic traits that would expected to minimize the effects of these constraints
How do organisms generate sufficient chemical diversity to enhance their chances of finding the rare beneficial chemical? How do organisms retain the capacity to generate so much chemical diversity when individual compounds or pathways become redundant?. The Screening Hypothesis proposes that certain metabolic traits (matrix pathways, non-enzymic transformations, branched pathways, shared pathways and enzymes with a broad substrate tolerance) would all help increase generation and retention of chemical diversity.

Some consequences of the hypothesis

Where can I read more about these ideas?

The following papers provide a more detailed consideration of the Screening Hypothesis and its implications.

1. Jones CG and Firn RD (1991) Phil Trans Royal Soc 333, 273-280.


A common-sense evolutionary scenario predicts that well-defended plants should have a moderate diversity of secondary compounds with high biological activity. We contend that plants actually contain a very high diversity of mostly inactive compounds. These patterns result because compounds arising via mutation have an inherently low probability of possessing any biological activity. Only those plants that make a lot of compounds will be well defended because only high diversity confers a reasonable probability of producing active compounds. Inactive compounds are retained, not eliminated, because they increase the probability of producing new active compounds. Plants should therefore have predictable metabolic traits maximising secondary chemical diversity while minimising cost. Our hypothesis has important implications to the study of the evolution of plant defence.

2. Firn RD and Jones CG (1996) In "Phytochemical Diversity and Redundancy in Ecological Interactions". ed Romeo et al., Plenum Press, NY. pp. 295-312.


The Screening Hypothesis is a model for the evolution of secondary product diversity that is based on extensive evidence, accumulated over several decades, which has shown that specific, high-potency biological activity is a rare property for a molecule to possess. Evidence to support this observation comes from studies conducted at the organism, organ, tissue, cell and molecular levels of organisation.

The proportion of naturally occurring molecules showing any biological activity will depend on the concentration at which they are assayed. Consequently it is argued that the meaningful comparative discussion of biological activity of any collection of endogenous compounds can only be made with reference to the concentration that will exist at the site of action of the compounds. When biological activity is only shown by a substance when assayed at concentrations above those that will occur in natural situations, it may be of pharmacological interest but such activity should not be assumed to be meaningful in terms of the discussing the possible role(s) that the substance might play in the organism that produces it.

The Screening Hypothesis is based on concepts which place very severe constraints on the evolutionary framework governing the generation, retention and use of secondary products. It is suggested that these constraints operate at a high level and that mechanisms to generate and retain chemical diversity must have evolved at an early stage in the evolution of life. Indeed it seems likely that secondary products may have been produced as a consequence of "biochemical inventiveness"which must have been inevitable when primary metabolism was evolving. "Secondary metabolism" might well have evolved concurrently with primary metabolism and primitive organisms may have evolved their "secondary metabolism" to exploit the inevitability of the generation of some chemical diversity.

Finally we return to the theme of the book, redundancy. It is apparent that a debate as to whether any secondary product is "redundant" becomes a semantic argument. If one defines redundant as no longer playing a role (i.e. the biological activity of a compound, which was once of value to the producer, is no longer exploited), the Screening Hypothesis would suggest that the many secondary products never had a role which resulted from their inherent biological activity hence cannot be considered redundant in this sense of the word. Likewise if one defines redundant as being superfluous or over-copious (Geddie W, 1959), the Screening Hypothesis would provide an arguments against the view that the majority of secondary products evolved as a result of any superfluous production hence most secondary products cannot be considered redundant in this sense. The Screening Hypothesis suggests that the term "redundant" would really only be appropriate in a very few circumstances. The Hypothesis suggests that the synthesis of many compounds which bring no short term benefit to the producer is a necessary part of the overall mechanisms employed by plants and microbes to produce the occassional chemical which possesses useful biological activity. Most secondary products are only "redundant" in the way that most antibodies are "redundant". The production of the majority of these substances results in no short-term benefit but short-term costs are compensated for by the longer term benefits that result when the rare biologically active compound is made. The majority of secondary products and antibodies are not redundant in any commonly accepted use of the word but are a necessary consequence of the need to generate chemical diversity.

3. R.D.Firn and C.G. Jones (1998) Avenues of discovery in bioprospecting. Nature 393, 617.

This is a response to a previous Briefing in Nature (392, 535) which considered the prospects for bioprospecting. We attempted to expain why bioprospecting has been so unsuccessful - biological activity is a rare property for a molecule to possess and contrary to common belief, natural products possess a high potency biological activity are rare. We suggest that combinatorial biochemistry could be a more productive way of exploiting genetic diversity.

4. R.D.Firn and C.G. Jones (1999) Secondary metabolism and GMOs. Nature 400, 13-14

Responds to previous arguments that GMOs can be shown to be non-toxic by testing or by metabolite screening. We argued that "the rules" of secondary metabolism will make it hard to predict the outcome of the genetic manipulation of secondary metabolism. Metabolite screening will be difficult because of the unknown nature of the possible new products and the fact that a full analysis of minor compounds in rarely attenpted. There is also a problem in knowing how secondary metabolism will respond to challenges - which induced profile do you test?

5. R.D Firn and C.G. Jones (2000) The evolution of secondary metabolism - a unifying model. Molecular Microbiology 37, 989-994.

Why do microbes make secondary products? That question has been the subject of intense debate for many decades. There are two extreme opinions. Some argue that most secondary metabolites play no role in increasing the fitness of an organism. The opposite view, now widely held, is that every secondary metabolite is made because it possesses (or did posses at some stage in evolution) a biological activity that endows the producer with increased fitness. These opposing views can be reconciled by recognising that, due to the principles governing molecular interactions, potent biological activity is a rare property for any molecule to possess. Consequently, in order for an organism to evolve the rare potent, biologically active molecule, a great many chemical structures have to be generated, most of which will possess no useful biological activity. Thus the two sides of the debate about the role and evolution of secondary metabolism can be accommodated within the view that the possession of secondary metabolism can enhance fitness, but that many products of secondary metabolism will not enhance the fitness of the producer. It is proposed that secondary metabolism will have evolved such that traits that optimise the production and retention of chemical diversity at minimum cost will have been selected. Evidence exists for some these predicted traits. Opportunities now exist to exploit these unique properties of secondary metabolism to enhance secondary product diversity and to devise new strategies for biotransformation and bioremediation

6. R.D Firn (2003) Bioprospecting - why is it so unrewarding? Biodiversity and Conservation 12,: 207-216.

Some economic analyses have placed high values on the chemical diversity residing in threatened habitats (Principe, 1996; Balick and Mendelsohn, 1992; Rausser and Small, 2000). In particular, bioprospecting (searching for new biologically active chemicals in organisms) is considered by some of its advocates to be a route to funding the preservation of biodiversity, especially in the LDCs. However, the large multinational pharmaceutical and agrochemical companies devote very little of their research effort to bioprospecting (Cordell, 2000). Why is this? The answer lies in the fact that any chemical (whether synthetic or natural product) has a very low chance of possessing useful biological activity. One model for the evolution of natural product diversity that is gaining increasing support suggests that the common belief, that natural products have been selected by their producers such that only biologically active natural products are commonly found, is wrong. Given that a random collection of synthetic or natural products have similar chances of containing a biologically active chemical and that synthetic chemicals are nearly always easier to synthesise on an industrial scale, it is to be expected that bioprospecting does not excite the multinationals. Although Rausser and Small (2000) argued that scientific advances will make bioprospecting more cost effective in future, an alternative scenario is presented where current biotechnological developments will further erode the value of bioprospecting. Hence there should be no expectation that large income streams will be available from bioprospecting agreements to help fund the preservation of biodiversity.

7. R.D. Firn and C.G Jones (2003). Natural products - a simple model to explain chemical diversity. Natural Products Reports, 20, 382-391.

A simple evolutionary model is presented which explains why organisms produce so many natural products, why so many have low biological activity, why enzymes involved in natural product synthesis have the properties they do and why natural product metabolism is shaped as it is.

8. R.D. Firn & C.G.Jones (2004) The evolution of plant biochemistry and the implications for physiology. In "The Evolution of Plant Physiology". Ed.

Some biochemical pathways are common to most plants ("primary metabolism"). Other pathways are widely found but are not universal (e.g. some parts of the carotenoid or lipid biosynthetic pathways). There are also minor branches leading from these pathways which are found only in a very small number of species ("secondary metabolism"). Why is plant metabolism like this? The Jones-Firn model to explain why plants and microbes produce so many "secondary metabolites" is based on the fact that the chances of any molecule possessing potent biological activity is very low. This model has now been extended by considering what other properties new molecules could bring to their producer. Two new propery classes are defined and it is proposed that selection would shape metabolism depending on the type of property. Furthermore, because particular properties will be associated with individual molecules, and not with particular pathways, it is predictable that many pathways will be multifunctional. By considering the evolution of regulatory systems controlling such pathways it becomes clear that such regulatory systems will have been shaped by the underlying multifunctionality of some pathways  with the result that "cross talk" is inevitable. An appreciation of these constraints may help those seeking to understand any physiological process that involve biologically active molecules (for example plant hormones or compounds involved in a plant's response to insects, fungi or bacteria).

The critics of the Screening Hypothesis

The Screening Hypothesis has been criticised and the following work should be consulted to read these criticisms.
1. Berenbaum MR & Zangerl AR (1996) Phytochemical diversity - adaptation or random variation. In "Phytochemical Diversity and Redundancy in Ecological Interactions". ed Romeo et al., Plenum Press, NY. pp. 1-24.
However, it should be made clear to readers that much of the evidence that is presented by Berenbaum & Zangerl concerns the furanocoumarins which are compounds which seem to bring about their major biological effects by covalently interacting with molecules. Making chemically reactive molecules as a defensive system will give rise to broad spectrum activity and will result in any molecules sharing the basic reactive parts of the structure being "active". However, this broad spectrum activity is hard to target, indeed such chemicals have a high chance of inducing a loss of fitness in the producer and "resistance" costs to the producer maybe significant.
Evidence from studies of compounds that react covalently with their targets does not contradict the basic ideas behind the Screening Hypothesis which is based on the much commoner reversible binding of ligands to proteins. We would argue that the furanocoumarins are a special case and it will be interesting to see how common such colvalent interactions are.
Readers might be interested in noting that "site-directed irreversible inhibitors" were once considered to offer very great potential as drugs (a book by G Baker in the mid 1960s??) but the approach did not yield the breakthroughs expected. Did the same happen in organisms?

Please engage us in debate. The hypothesis is there to argued about and we really do enjoy thinking about its strengths and weaknesses. We have some more papers in preparation which aim to develop the ideas further. We are assembling evidence which is consistent with the predictions we made in the original article consequently we would welcome being made aware of evidence which supports or contradicts the predictions. We would also welcome evidence as to the frequency of biological activity in collections of chemicals. There must be masses of data in the files of screening companies which would be valuable - we don't need to know the identity of the copounds or even the exact screen being used!
Richard D Firn,

Feb 2004