FW: Mice and elephants

["Wood, Michael" <woodm@aecl.ca>]

The following is the response to Bernie's "Mice and elephant" post from
one of our senior Scientist.  As Paul is not on the list I am posting it
for him.

Mike Wood
Radiation Biology and Health Physics Branch
Chalk River Laboratories
Woodm@aecl.ca


>Practically no one expects one energy deposition event in one DNA molecule to
>lead to cancer in either mice or elephants.  In fact, there is a reduction of
>10^18 between the initial numbers of hits in DNA and the final probability of
>a cancer even if all humans, for example, are held to be genetically
>identical, which is absurd as the natural range of radiation resistance in
>human populations is at least 7 fold.
>
>Mice in the wild have a half-life of about two weeks, so events which happen
>in three years time are irrelevant - they are not under selective control.
>So for a mouse, the possibility that a particular energy deposition event
>leads to cancer, in nature, is irrelevant.  A difference in the probability
>of repair of 10 or 100 fold (from 10^18 to 10^16) is irrelevant.  The number
>of cell doublings separating a 30 gm mouse from a 2 tonne elephant is on the
>order of 15-16.  For mice, there is less selective pressure to evolve high
>fidelity control and repair systems, even though inbred strains in the lab in
>fact show good controls, but for the elephant there is.  Getting to
>10^11-10^12 cells successfully in the case of the mouse doesn't require the
>same quality control that getting to 10^15-10^16 cells does in the case of
>the elephant.
>
>For an elephant, the problem is simply to last the full 75 - 100 years in the
>wild before dying.  Increasing the efficiency of DNA repair, or of
>immunological cell surveillance by some amount makes this possible.  But even
>for an elephant, the only selective advantage comes by pressure before and
>during the age of reproduction.
>
>Small children are more sensitive because they are still growing, and clonal
>events, which may arise at random or from energy deposition, have more cell
>doublings in which to play out.  So a cell which loses a cancer repressor
>gene has a better chance in a child of leading to a cancer.  Furthermore, the
>only interesting thing about ionizing radiation as a carcinogen is that it is
>entirely stochastic as to when DNA damage can occur in the cell cycle, and
>some cells are always in a high risk part of the cell cycle.
>
>But in retinoblastoma, for example, cancers are present at birth, and do not
>reflect ionizing radiation so much as simple chemical noise in systems which
>use DNA for an informational molecule.  Shit happens.  
>
>So I have some questions for Bernie:
>
>What sort of biology is it that doesn't take into account genetics,
>development, and known aspects of the cancer risk process?  What sort of
>biology is it that doesn't take into account the genetic uniqueness of each
>sexually reproduced organism on earth?  What needs to be done to get
>radiation physicists to think like biologists?  Is there any hope?
>
>Paul Unrau
>RB & HP Branch
>AECL
>UnrauP@aecl.ca
>
>
>
>
>
>
>