Why would cancer developing in a stem cell in the basal layer of the
skin's epidermis be potentially more harmful than cancer arising in a
specialized cell above it?

The three main skin cancers are basal cell carcinoma, squamous cell
carcinoma, and melanoma.  By far the worst is melanoma.  Melanocytes
reside in the layer at the bottom of the epidermis.  It is not clear
whether melanomas arise from fully commited melanocytes, or less
differentiated melanoblasts, or both.

So - I think the answer is - underlying cell type is more important
than level of differentiation.  Beyond that, with melanomas anyway, no
distinction has been drawn based on the degree of differentiation of
the initial aberrant cell.

Hello Tammy,

I'm not sure about the validity of the statement; in what context have
you come across it? Basal cell carcinoma is freqently slow-growing,
rarely metastisizes and, with treatment, usually has a good prognosis.

I think we have to be careful to distinguish between basal
keratinocytes (that may lead to basal cell carcinoma), and other cell
types that may exist in the epidermal stratum basale, notably the
melanocyte, but possibly including precursor cells of both types.

My take on the question is that it seeks clarification of the
increased severity of tumours arising from less differentiated cells
in the basal layer, not necessarily keratinocytes, versus tumours
arising from fully differentiated cells.  There may be some marker
studies where there is expression of "less differentiated" proteins,
but it is never clear whether this is due to the tumour arising from a
precursor cell, or due to aberrant expression in a fully
differentiated cell.

Tricky business this tumorigenesis stuff.

Less differentiated tumours are generally more aggressive than
better-differentiated ones.  However, is that due to the extent that
dedifferentiation has occurred rather than to the original level of
differentiation of the initial tumour cell?  I was always taught that
it is the extent to which the tumour cell has become different from
its original self that determines aggressiveness, most particularly
with respect to the *change* in rate of division and to the way it
responds to normal growth-regulatory mechanisms.  As far as I know,
adult skin stem cells are highly responsive to growth regulation. Are
they susceptible to dedifferentiation and is this likely to make them
more "embryonic-like" and therefore more likely to go out of control? 
The evidence seems to be against this, since basal cell carcinomas,
which arise from one type of precursor cell in the skin, are not as
aggressive as some other skin cancers.  The aggressiveness of melanoma
has been attributed to the migratory characteristics of melanocytes,
rather than to their original level of differentiation.

The trouble is that we do not know from where this question has
arisen.  Does it expect a hypothesis to be developed from a set of
facts that have been presented, in a lecture for example?  If yes, the
hypothesis might or might not reflect reality.

My, doubtless imperfect, understanding is that the notion of
"dedifferentiation" is losing popularity.  Histologically, it may be a
useful concept, and as tehuti says, there is an association between
tumour aggressiveness and cell morphology.  However, at a molecular
level, it seems unlikely that the programme of differentiation that a
cell has followed can be somehow reversed.  Recent studies concerning
the partitioning or mRNA species facilitated by centrosomes suggests
that the process is in effect irreversible, as it should be.  Other
aspects of the differentiation program are implemented by
transcriptional silencing, either through the enhanced production of
transcriptional repressors (often creating positive feed-back), or
through gene methylation.
What can easily happen is that genetic changes, often MAJOR,  that can
be brought about by increased genomic instability in cancer cells, can
lead to the re-expression of proteins associated with earlier states
of differentiation.  Similarly, changes in phenotype or morphology can
occur that may echo those of other cells, including the ancestors of
the cell in question.

Int J Cancer 2003 Mar 1;103(6):709-16 
Matrix metalloproteinase-19 is expressed by proliferating epithelium
but disappears with neoplastic dedifferentiation.
Impola U, Toriseva M, Suomela S, Jeskanen L, Hieta N, Jahkola T,
Grenman R, Kahari VM, Saarialho-Kere U.

Cancer Lett 2003 Mar 20;192(1):83-7 
p130 expression in thyroid neoplasms: its linkage with tumor size and
dedifferentiation.
Ito Y, Yoshida H, Uruno T, Nakano K, Takamura Y, Miya A, Kobayashi K,
Yokozawa T, Matsuzuka F, Kuma K, Miyauchi A.


Curr Biol 2003 Jan 21;13(2):134-9 
Dedifferentiation of Primary Spermatocytes into Germ Cell Tumors in C.
elegans Lacking the Pumilio-like Protein PUF-8.
Subramaniam K, Seydoux G.

Hi tehuti!
Your first paper refers specifically to "histologic
dedifferentiation".
In the second, "dedifferentiation" does not appear in the paper
abstract, so it is hard to judge what they are saying or what
importance they place on it - short of getting the paper.
The third paper appears to relate to a knock-out experiment in C.
elegans.  The researchers have disrupted the function of a particular
gene and found that it plays a role in differentiation of this cell
type by blocking mitosis.  Take out the gene - the cells can again
divide - a sort of de-differentiation I guess.  If I overexpress ras
to cause a similar effect, is this dedifferentiation?

The crux of the question posed was whether tumours arising from stem
cells in the skin were more likely to be aggressive (as I understand
it).  I'm not sure if this is known with any certainty. 
Dedifferentiation, whether it is a bona fide phenomenon or not, is
probably not an issue here.

In database indexing and searching, the appearance of a word in a
title is normally taken to be of greatest significance in judging the
"aboutness" of a paper (setting aside titles which try to be funny or
too clever).  I was merely trying to point out that the term
dedifferentiation retains sufficient clarity of meaning to be used in
this age.

I agree the question is about the potential aggressiveness of tumours
arising from stem cells.  However, the nearest real example we have of
this is the basal cell carcinoma, which is not particularly
aggressive.  That is why I wonder if this question deals with real or
hypothetical situations.

stem cells divide to give daughter cells.
 When they divide,one remains at the basal layer and continues as a
stem cell. The other goes on to specialise-moving through the 5
epidermal layers to the outside.
So a cancer mutation in a stem cell will be passed on to all its
'daughters' and groups of cancer cells lead to tumours. A mutation in
a specialised cell(i.e. not a stem cell) will be lost with the skin
cell when it sloughs from the skins surface.

What amanda1 says may be true for keratinocytes, as they differentiate
to squamous cells and ultimately are lost.  However,  the mutation
that occurs may interfere with the differentiation and migration
process.  Melanocytes, however, remain in the stratum basale, at the
interface between the epidermis and the dermis.  Tumours of
melanocytic lineage are far more aggressive, and have a much poorer
prognosis if not treated early, than other skin cancers.