Williams syndrome = negroid - gene(s) ?

Tim Raine tr223 at hermes.cam.ac.uk
Sun Nov 5 14:13:36 EST 2000

> From chromosomes to CAG repeats, this business of genetic repeats seem to
> feature in neurodegenerative conditions, though here in completely
> ways. Huntington's must have over 40(? - varying refs here) CAG repeats
> this sets up some nasty processes.
> One ref I have suggests that beyond a certain point the protein no longer
> becomes soluble, thus cannot be folded by heat shock proteins and so
> inactive but cannot be set up for proteosomal degradation, thus allowing
> aggregates ... . Matches with reports of heat shock protein 27 being
> implicated in Parkinson's disease, and possibly MS or Alzheimer's (mem
> failure - amyloid?!). Recently, research suggested that the fibrilles of
> Alzheimers like that of prions ... . Let's go to McDonald's .. Is there
> anyone here who can shed light on why polyglutamine repeats can be so
> dangerous for neurons?
> How could someone possibly equate William's with negroid features? Somehow
> Kofi Anan does not appear to be someone lacking in spatial abilities or
> You won't find a single William's with that level of ability and skill,
> there are plenty of Kofi's in the world.
> Wandering,
> John H.
> Remove 4X

Dear John,
Though I'm reluctant to further a thread that started with such offensive
and absurd a post, your question is an interesting one and one in which I
have a little interest myself.

With regard to why CAG repeats lead to aggregates, Max Perutz showed that a
polar zipper structure was formed which was clearly capable of driving

Histological studies show that the appearance of inclusion bodies (IBs)
occurs prior to the onset of disease symptoms, so it is reasonable to
suspect that the IBs themselves may be responsible for pathology.
Greenberg suggested that ubiquitination of the mutant proteins (driven by a
failure of normal expression) drove their aggregation via mutual targeting
to the proteosome.  By inhibiting ubiquitination, Greenberg showed that the
number of aggregates dropped, but that the numbers of apoptotic cells also
increased.  However, by inhibiting ubiquitination, Greenberg did rather mess
with essential cell biochemistry and also struck at a rather arbitrary point
on the proposed aggregate - cell death pathway, so I'm not sure how much we
can draw from his results.

Likewise, Cummings and Zoghbi last year crossed transgenic mice with the
E6-AP enzyme (essential for the last step in ubiquitination) knocked-out
with mice transgenic for the trinucleotide repeat expansion mutant SCA-1
gene (which leads to spino-cerebellar ataxia).  The progeny showed decreased
numbers of aggregates, but increased pathology.

Again, Zoghbi generated mice transgenic for the SCA1 mutant, but with
Perutz's 'self-association' domain cut out.  This time, no aggregates
formed, but the mice still showed the SCA pathology.  (Perutz came back at
this one arguing that deletion of the domain would mess up protein function
and lead to pathology anyway and criticised her lack of controls).  Zoghbi
herself has reported that her mice did not show such severe pathology, so
once again, the focus is returning to the role of the aggregates.

So we have 3 opposing views:
1) Non-toxic proteins form toxic aggregates leading to cell death.
2) The mutant proteins themselves are inherently toxic, but their aggregates
are non-toxic and hence afford some degree of protection (but not enough)
hence the increased pathology in Cummings + Zoghbi and Greenberg's studies.
3) That the mutant proteins lead to death and the aggregates are just an
epiphenomenon with no consequence at all!

Studies done using GFP tagged Huntindin mutants showed that the mutant
proteins seemed to recruit normal forms of other poly-glutamate containing
protein into the aggregates, acting as 'seeds' (this is where your analogies
with prion theory come in).  These proteins include CREB-binding protein
(CBP) and TAF2150 (another CREB coactivator).  Both of these proteins have
been isolated from the aggregates formed in mouse models.  Subsequently,
researchers have shown that overexpression of either CBP or TAF2150 in cells
containing the mutant polyQ protein help to rescue the cells from apoptosis,
thus suggesting that the aggregates may be toxic (model 1) by virtue of
their sequestration of other vital nuclear proteins.

If we are to treat these diseases, we need to know how the cell attempts to
handle the mutant proteins.  There are two possible pathways:
1) via the 26s proteosome pathway
2) via chaperone pathways
Markers for elements of both of these pathways colocalise with the
aggregates during confocal microscopy visualisation.
Evidence for (1): Inhibition of the proteosome increases aggregate
Evidence for (2): Hsp70 and Hsp 40 are seen sequestered into the aggregates
very early on indeed.  Overexpression of HDJ-2 lead to increased aggregation
in some cells lines but decreased in others, suggesting a cell type specific
mechanism.  Studies using the GRO-EL and GRO-ES system of bacteria
demonstrate that transfecting the sequence coding for even a single domain
of either protein into cell lines containing the mutant protein will
effectively reduce aggregate formation with a corresponding drop in

So, cells with aggregates do seem to show increased rates of apoptosis,
whilst transfection of proteins capable of reducing aggregate formation
reduce apoptosis.  To say anything more definitive will take a few more
years of research!

I hope this answers some of your questions.  If you want to read a really,
really excellent review on this then I'd recommend a review published by
Cummings and Zoghbi in the Annual Review of Genomics and Human Genetics
(Vol. 1; 281-328).  It discusses all of this in much more detail!

Best wishes,

More information about the Neur-sci mailing list