Hi. I must admit I haven’t read the whole thread in excruciating detail, and I have only paged through jorolat’s homepage.
I have a number of issues however:
My first grad school rotation was with Susan Rosenberg, who studies stationary phase mutation. To generate stationary phase mutations, she grows E. coli that cannot digest lactose (lac[sup]-[/sup]) to saturation, and then plates on a plate with lactose as the only carbon source. Regular bacteria can digest lactose, but these ones have a +1 frameshift in one of the genes responsible for lactose digestion.
In stationary phase, she sees a population who revert the +1 frameshift and become lac[sup]+[/sup]. The mechanism is complex, but basically deals with a global downregulation of proofreading and upregulation of error-prone polymerases which are active during DNA repair. This causes slippage and high rates of mutations at mononucleotide repeats.
A couple of things:
- Stationary phase mutation hasn’t been seen in anything except bacteria (see below).
- There is no targeting to the specific non-functioning lac gene.
- The bacteria become globally mutagenesis prone (mut[sup]+[/sup]).
- This occurs in a population and only a few out of every million bugs become lac[sup]+[/sup]. These bugs get a huge advantage over their lac[sup]-[/sup] neighbors.
So no stationary phase mutation in anything but bacteria. But, a similar phenomenon has been noticed in cancer cells and in yeast. A gene responsible for chemotherapeutic drug metabolism or transport can become amplified by tandem copying (often millions of times, creating visible chromosome structures called double minutes) in order to lend resistance. Basically, if one copy of the gene can give 1/1000 resistance, 1000 copies will give full resistance. Problems again:
- This occurs in a population.
- Cancer cells have often turned off cell cycle checkpoints that would cause an ordinary cell to commit suicide if it tried to amplify its DNA.
- A cell who amplifies its resistance gene twice will have a two-fold advantage over its neighbor. Such a process is kind of like a snowball – a tandem repeat can expand easier than a non-repeat, so amplification four fold from two fold is easier than from one fold to two fold.
The problem with jorolat’s model (if I understand it) is that there is no population of cells here. Amplified or stationary phase cells are clonal – all descending from one cell. This doesn’t happen in the germ line (all the kids would look the same!) or in the cerebellum. It has never been described. Also, stationary phase mutation and amplification are non-directed. While this was unknown in 1988, it is very clear now. This is not an example of the environment feeding back into the genome.
Next, there are examples of the environment feeding into genetic regulation. These are not direct, however, although they may involve inherited change. These mechanisms are epigenetic regulations like methylation, imprinting, heterochromatin structuring by histone acetylation, and many other modifications to the DNA structure. These do not (that I know of) ever change the genetic information, but may silence or enhance gene expression in generations. Certain genes are turned off by the maternal or paternal germ line cells in imprinting. This can lead to certain diseases – Angelman syndrome, Prater-Willi syndrome, Beckwidth-Wiedemann, and certain other disease which can occur in a non-Mendelian form.
If I were looking for instances of Lamarckian evolution, I would focus on epigenetic inheritance. Stationary phase mutation and amplifcation are poor candidates, IMHO.
Also, I’d read up a little about Lysenko and the havoc he wrought in the Soviet Union when he convinced Stalin that Lamarck’s theories were more Marxist and must therefore be correct.
http://www.dcu.ie/~comms/hsheehan/lysenko.htm
http://www.princeton.edu/~browning/215/readings/gould.html (from Steven Jay Gould’s book The Panda’s Thumb