I have just read John Woods latest attempt to kick-start this newsgroup,
where he put forward some criticismsof Kacser & Acerenza's latest paper
`Universal method for achieving increases in metabolite production'
[Eur. J. Biochem. 216:361-367 (1993)]. While I am all for the critical
evaluation of published material, I think he should supply us with some
reasoning behind his somewhat general statements.
In reply to some of his comments I would like to say that `theoretically'
this method is universal. Any metabolic system in which we can consider
suitable for metabolic control analysis can be dealt with by this
method. This is more than can be said about mutation/selection. When was
the last time you read a paper on the improvement of milk production in
cows following a round of muation and selection? As John points out,
the method is closely analagous to the branch point relationships,
except we are dealing with finite changes and hence can ignore the
control coefficient aspect of the relationships. The computational
aspect of the method, i.e. chosing which steps which require to be
modified and by how much, given the constraints that we want to keep
most o fthe rest of metabolism constant, can be obtained using the Null
space of the stochiometry matrix as for the branch point relationships.
By removing the columns from the matrix representing the steps which you
dont want to change, it is possible to obtain the solution relatively simply.
Any problems with the method will come in application. As John points
out, (or rather Herbert Sauro (or perhaps someone told him as well)),
bi-molecular reactions pose a problem in application as they extend the
number of steps which have to be manipulated. They do not invalidate the
method however. A more serious experimental problem would be carrying
out the enzyme manipulations in the exact ratios required. The control
and manipulation of different plasmids at different copy numbers will
have to be realisible for this method to succeed. This may not be
possible now, but who knows about the future (or what goes on inside
biotechnology companies?).
To finish my ramblings I would like to make the following points:
1) If mutation and selection (and to be fair to milk production,
artificial selection) are so good, why are the biotechnologists
attempting to increase metabolite production using genetic manipulation?
2) If genetic manipulation is going to be used, it is better to have
some ideal to aim for than scrambling about in the dark. Even if this
aim is not wholly practically acheivible, partial success will probably be
much more succesfull than manipulating single steps in isolation.
3) The method is such that even though some steps will theoretically
have to be manipulated, because of the step wise reduction in the
factors by which steps upstream from a branch have to be changed, many
of these steps can be left as normal, without causing to much problems.
4) This method is very simple to understand and does not involve any
fancy maths. Hence the theoreticians amongst us might feel that it is
not worth considering. Its simplicity however may be very attractive to
the experimentalists. Dont dismiss it too lightly, John.
Rankin Small (unemployed)
