Model of an Internal Evolutionary Mechanism linking "Stationary-Phase Mutations" to the "Baldwin Effect" Part 2: The Bacterial Model  ©1999 John Latter

Contents:

1) Introduction
2) The Lac Operon
3) Origin of the "Directed Mutation" hypothesis
4) "Stationary-Phase Mutations"
5) A single Homeostatic Mechanism
6) Conclusion

1) Introduction

   Of initial interest is the actual phenomena that Cairns was investigating, why he proposed the hypothesis of "Directed Mutation", and how subsequent experiments by other scientists have led to it's modification.

   The main section then briefly reviews a number of experiments that have exhibited the phenomena and suggests that a single mechanism, triggered by varying degrees of disruption to homeostasis, may be responsible for the diversity of results that have been observed. This is a "stand-alone" argument and the AONE is not referred to until the next section where a comparison is made with post-notochord model.

   A technique used in the experiments referred to is to engineer a mutation in a strain of bacteria which then necessitates their being supplied with the "product" they have become dysfunctional for (there are various types of mutation and a concise description of how a "point mutation" is engineered can be found at reference [3]). An experiment would then begin, for example, when the product was removed.

   Some of the experiments involved mutations to the "lac operon" and a simplified description of its operation appears first.

2)The Lac Operon

  The "operon" model of gene regulation was first proposed by Monod and Jacob in the 1960's and there are two general types: inducible and repressible. Figure 1 shows the basic configuration (omitting the cAMP/CAP complex) of the inducible operon for the sugar lactose

	The repressor molecules are continuously being produced and in the absense of lactose derived substrate molecules are able to bind to the "operator" site of the operon.     This prevents the RNA polymerase that recognises and binds to the "promotor" site from moving along and transcribing the structural genes (the "OFF" state).
	When the substrate molecules are present (figure 2) they bind with the repressor molecules and by changing their shape prevent them from occupying the "operator" site.     RNA polymerase is now able to move along from the "promotor" site and transcribe the structural genes. Subsequent translation produces the enzymes necessary to break down the lactose into glucose and galactose (the gal operon eventually converts the galactose into glucose). The "ON" state.

   Repressible operons that are normally "ON", such as those specifying particular amino acids, can be switched "OFF" by a homeostatic feedback loop. If, for example, a bacterium is supplied with a particular amino acid then the amino acid itself (or "end-product") binds to the associated co-repressor molecules and by changing their shape enable them to occupy the "operator" site on the gene. Transcription of the gene is thus inhibited until concentrations of the amino acid fall back to more optimum levels.

3) Origin of the "Directed Mutation" hypothesis

   An experiment representative of those that Cairns conducted was performed on a strain of bacteria (e. coli) that were unable to utilize the sugar lactose [4]. First the bacteria were deprived of all sugars for a few days, during which time they entered a starvation or "stationary-phase" state and ceased replicating, and then they were provided with lactose as the only source of nutrition.

   Once the lactose was added the number of bacteria that subsequently mutated enabling them to utilize it while not replicating far exceeded the number that conventional theory could account for. It was this appearance of the mutations being specifically directed to the presence of the sugar that contributed to the forming of the "Directed Mutation" hypothesis.

4) "Stationary-Phase Mutations"

   Seeking to disprove the hypothesis Mittler & Lenski [5][6], and separately, Sniegowski [6] performed their own experiments and each concluded that the mutations observed while their bacteria were in the stationary-phase state weren't "directed" at all but were instead attributable to the "non-specific stresses of starvation" (my italics).

   The two points of significance concerning these particular experiments are i) the strains of bacteria used had different mutations to those employed by Cairns and ii) the number of mutations still exceeded what conventional theory could account for (the number of observed mutations varies, in the appended references, from between 100 to 100 million times greater than expected).

   Taking these and the results of other experiments into account has led to the phenomena becoming more commonly known as "Stationary-Phase Mutations" (now "Adaptive Mutations"). The change in terminology has also lessened controversy although in 1997 a researcher in this field continued to describe it as "the mutation whose name one dare not speak" [7].

5) A single Homeostatic Mechanism

    Homeostasis is sometimes described as "self-regulation" which is a phrase with the potential to evoke similar emotional responses as those caused by "directed mutation". A more objective point of view can be obtained by seeing homeostasis as, for example, the dynamic maintaining of equilibrium varying about a mean within a biological system.

    If consideration is given to the possibility that stationary-phase mutations in bacteria may be part of a homeostatic function then this would arguably be a lesser capability than that of replication itself [note 1]. After briefly reviewing, from a homeostatic perspective, various experiments that have produced stationary-phase mutations the Conclusion gives a general indication of the homeostatic range within which they could occur before considering an obvious objection: "If homeostasis is involved then why, in these experiments, didn`t more of the bacteria revert?".

   In general: bacteria are naturally integrated organisms and depending on which part of the genome is artificially mutated will depend the effect on internal equilibrium, as much as selected function, once the compensatory nutrients are removed. For example, if a mutation is engineered towards the base of an integrated homeostatic hierarchy then subsequent reversion of this mutation would be localized, and from an external perspective, may give the appearance of being "directed".

   On the other hand another mutation occurring at a higher homeostatic level, or at the same level but having a greater effect, may cause restoration of overall internal equilibrium to take natural precedence over (any) restoration of selected function. In these instances the reversions would initially appear less directed and consequently attributable to "the non-specific stresses of starvation" (i.e. disequilibrium).

   Specific experiments indicating restoration of equilibrium may be involved in stationary-phase mutations include the following:

a) An experiment performed by John Cairns where an original flaw in the lactose enzyme gene wasn`t corrected but compensated for by the way the cell's protein synthesis machinery read the gene [4], i.e. it didn`t matter how the "upset" to equilibrium was restored just as long as it was within the range that it could be.

b) Barry Hall performed an experiment on bacteria that were auxotrophic for the two amino acids tryptophan and cysteine (i.e. they had to be supplied with the appropriate products) and hence designated Trp-Cys-. Bacteria then starved for tryptophan (but not cysteine) reverted to Trp+ (but not Cys+) while those starved for cysteine (but not tryptophan) reverted to Cys+ (but not Trp+) [5].

   Superficially these results appear to support the hypothesis of Directed Mutation but it is suggested that only when one or other of the compensatory nutrients were removed did the "breaks" in the respective (but integrated) homeostatic loops affect equilibrium to the extent that reversions occurred .

c) In one experiment (Cairns) lac- bacteria were provided with "IPTG" [5] which is a gratuitous inducer in that it causes transcription of the lac operon but thereafter cannot be metabolized by beta-galactosidase. No revertants were detected which, from a homeostatic viewpoint, would be explained  by there being no subsequent energy available to the cell, i.e. without an "end product" there wasn`t anything to re-establish equilibrium with. An open-ended loop.

d) In some experiments mutations have been found in genes additional to the one selected for [4][8]. A homeostatic explanation would be that the bacteria re-established equilibrium within laboratory environments different from those where any previous stationary-phase mutations had contributed to evolution of the pre-engineered (wild) strains. A ripple effect, analogous to water finding it`s own level.

e) Other results indicate that homeostasis may be involved, e.g. experiments have been performed requiring double reversions and in one of these [5] excision of an insertion sequence preceded the enabling of transcription. This particular experiment was performed by Barry Hall and is interesting for two reasons: a) Intuition suggests that an engineered mutation consisting of an insertion sequence would generally have a greater effect on equilibrium than a point mutation. b) Supporting this is the fact that Hall reported excision of the insertion sequence in growing cells is so rare that he couldn't find it.

6) Conclusion

   Perhaps the most important factor to be considered in answer to "why didn`t more bacteria revert?" is contained in a statement by John Cairns: "If you want to know whether creatures can turn on the mutations they need it`s not fair to kill them, you want to know if they can turn on a mutation by worrying about the environment ".

   At one end of the spectrum a single engineered mutation on the bacterial genome may not affect internal equilibrium sufficiently to cross the lower threshold at which stationary-phase mutations begin to occur. At the other extreme a mutation on another part of the genome may affect equilibrium to such an extent it is beyond the upper limit of the "range of recovery" and the bacteria die.

   Though common limiting factors are almost certainly involved any experiment tending to affect equilibrium towards the upper limit would produce a smaller number of revertants than another experiment that, while upsetting equilibrium enough to cause stationary-phase mutations, was more within the range of recovery.

   If stationary-phase mutations are part of a homeostatic function then the foregoing also indicates how it would contribute to maintaining integrity of the bacterial genome: a random mutation that upset equilibrium within the range of recovery would tend to be reversed but not necessarily so, it may be "modified" and/or trigger the ripple effect.

Note 1: Only when the stationary-phase mutations had occured did those bacteria, in principle, regain the potential to replicate and it is from this perspective that restoration of equilibrium via the proposed mechanism is seen as a lesser capability than that of replication. [return to text]

[1] Cairns, Overbaugh and Miller (1988), The Origin of Mutants, Nature 335: 142-145 [Abstract: subscription required]
    http://www.nature.com

[2] Latter (1999), Directed Mutation and Galileo
     http://members.aol.com/jorolat/direct.html

[3] Site-directed Mutagenesis [Internet Archive]
     http://johns.largnet.uwo.ca/shine/cmhf/sitedrct.htm

[4] Richardson, Survival of the Mutable
     http://www.zinkle.com/p/articles/mi_m1511/is_n9_v15/ai_15770748

[5] Goodman, Directed Mutation: Heredity Made to Order [Internet Archive]
     http://www.aaas.org/spp/dspp/dbsr/EVOLUT/goodman.htm

[6] Jones, Suntoke, Shereck, Molecular Biology of Prokaryotes - Directed Mutation
    http://www.bio.cmu.edu/Courses/03441/TermPapers/96TermPapers/directed/index.html

[7] Rosenberg (1997), Mutation for Survival, Genetics and Developement Vol 7, No. 6 [PubMed]
     http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dopt=Abstract&list_uids=9468794

[8] Beardsley (1997), Beardsley (1997), Evolution Evolving, Scientific American Sept 97
     http://www.dhushara.com/book/evol/evev.htm