After William Bains wrote:
>Nature Bio/technology has had lots of letters saying, in essence, 'why did you publish this thing in
an industry newsletter?' One answer, obviously, is that it is a research
journal, not an industry
newsletter.
I asked:
>Where did these letters appear? Are they in the May issue (which we
don't have yet) as I don't find any mention of the Strohman article in
the April issue?
As we have now received the May issue of Nature Biotechnology and as he
has been too modest (or too busy) to answer my question I'll answer it
myself: on p. 396 there is a one-page Commentary by none other than
William Bains, under the title "Should you hire an epistemologist?" This
also is well worth a read (and is easier to read than Richard Strohman's
article on pp. 194-199 of the March issue), and has quite a lot of
interesting things to say. As a flavour of his point of view:
"Beadle and Tatum's 'One gene -- one enzyme' has become the industry's
'One gene -- one drug target' with unprecedented unanimity. Can we have
got it so wrong? I am sure we have, and am equally sure it does not
matter."
The reason it does not matter, he thinks, is that industry has always
muddled its way through so far, but the successes have been enough to
pay for the cock ups, and this is how it will be in the future. Maybe
true, but I still have some doubts.
It was in about 1970 that I first heard a lecturer from the drug
industry announcing that "yes, until now, it has been mainly a matter of
hit and miss with random screening of haphazardly selected compounds,
but in the future we'll do much better because we'll *design* the
drugs". A quarter of a century has passed, but it is still largely a
matter of hit and miss, and industry spokesmen continue to speak with
great confidence about how the latest trendy approach is going to make
it all more reliable.
The other point is that the cost of the cock ups in biotechnology over
the past two decades has been almost inconceivably huge. Buildings
built, salaries paid, chemicals wasted, experimental animals killed,
scientists out of the main stream put out to pasture, university
departments doing old-fashioned work closed down or merged with their
trendier neighbours... If all this is added up it is far from clear that
the benefits have been greater than the costs.
Let us put it in terms of an example that will be very familiar to
members of this group, but which we have not succeeded in conveying to
the great world outside.
Start with something that all biochemists know (or think they know):
Phosphofructokinase controls the glycolytic flux.
So, overexpressing it should result in increased glycolytic flux; right?
Wrong.
Heinisch [Mol. Gen. Genet. 202, 75-82 (1986)] tried this more than a
decade ago. He overexpressed phosphofructokinase 3.5-fold, and found *no
change* in the flux to ethanol.
So, there must be something odd about yeast; right?
Wrong.
Essentially the same experiments in potatoes gave essentially the same
result.
Similar experiments in numerous organisms (fruit flies, moulds,
bacteria, trypanosomes, mice... you name it) give similar results.
Yet a lot of the biotechnology industry continues to proceed on the
basis of an essentially wrong idea, and unless the implications of these
experiments begin to be taken seriously (and not dismissed as
philosophy) the next two decades of biotechnology will be as sterile as
the first two have been.
Maybe this won't matter. Maybe there will be just enough fruit to
convince the politicians and investors that they are getting something
for their money. But maybe not. What then?
Athel
