telomeres and aging
davidk at hgu.mrc.ac.uk
Thu Apr 18 08:54:22 EST 1996
Jean-Pierre Issa wrote:
> Finally, I would like to hear your thoughts on the implications of
> telomerase activity in mouse cells: Mice still senesce in-vivo and
> die... I am not sure that the cancer incidence in non-inbred mice is
> much higher than that seen in humans... Doesn't this argue against the
> telomere loss/aging connection and the telomerase/cancer connection ?
I saw the words 'mouse' and 'telomerase' in the same sentence so I feel I should reply, having
some practical interest in this subject :-)
This will sound like a telomere/cancer post, but in this field cancer and senescence are simply
the flip side of the same coin and data on one impacts and inputs to the other.
Yes, mice are weird.
1. Mice telomeres appear much longer (5-10 x) than human ones. This seems true for both
inbred mice and wild mice (ref: recent Garagna et al paper in Chromosoma on Rb mice).
2. Long telomeres don't seem to be necessary for mice to live. There is a US population of
Mus spretus in the lab with short (a la human) telomeres and they seem OK. NB the European
spretus stocks have long telomeres, like other mice [I'll be including this datum in a telomere
review later on this year]. The guess is that the US mice have picked up a mutation while
becoming partially inbred which has reset their telomeres to a smaller length; the European mice
are probably closer to the wild state, given we re-trap them every now and again :-)
3. Telomere loss rates cannot be measured normally, but in the strange US spretus Greider's
lab have reported a 50bp/division loss rate. Based on this and the size of telomeres in old mice
(pretty much like young mice) we would say that mouse telomere shortening occurs at a rate like
in humans, one consistent with incomplete replication losses. This predicts that there will be
only a small percentage change in telomere loss during the lifespan of a mouse, consistent with
the exptl data.
4. Mouse telomerase is present at low levels in primary mouse tissues but is strongly
up-regulated in mouse tumors. Exampel: data coming out from the de Lange lab reports
telomerase assays and telomere length determinations on a mouse tumor system, namely the
MMTV-driven Wnt-1 mammary tumors. Here normal tissue has much less telomerase than tumors but
the telomere lengths are indistinguishable. This is pretty difficult to explain if one is
suggesting that telomerase re-activation is a requirement for tumor formation because it prevents
There is another way to look at this. Replicative immortality is an "ancestral" function (i.e.
something our unicellular ancestors had). Replicative senescence, differentiation and so forth
are metazoan features superimposed on this basic system. Telomerase inhibition in somatic cells
may be part and parcel of the same germ/soma difference.
When one is talking about a cancer cell being replicatively immortal one might think of it as
returning closer to the ancestral state (immortal) - certainly cell fusion experiments indicate
that immortality is a recessive phenotype, in that senescence is something additionally imposed
on an otherwise immortal cell.
Maybe then when one sees a cancer cell it is inactivating a general pathway involved in this
extra layer of division control. One of the unavoidable consequences (by-products, artefacts,
whatever!) of this is that telomerase repression is removed as well. In other words, you cannot
make a cancer cell without also activating telomerase - not because you *need* to, but because
the pathway does not allow you to do differently.
So, what has all this got to do with replicative senescence? Well, it is not clear. The data is
reasonably strong to suggest some link between replicative senescence and telomeres in humans.
However, just because a mechanism acts in humans doesn't mean mice must do the same. The
unchanging telomere lengths in mice is not consistent with a telomere-senescence link, and one
could argue the telomerase up-regulation in mouse tumors is a by-product of (hypothetical) basic
cellular changes in cancer cells (reverting to the ancestral phenotype). Mouse cells may age in
a different way to human cells (lack of decent DNA repair or whatever) and I don't have a problem
with this. Orgamismal ageing is a very complex and diverse subjet, and there is no reason to
expect that all mammalian cells in vitro which show replicative senescence will do so for the
same reason. It would be nice if they did, but Nature is usually never so nice to us :-)
More information about the Ageing