Page 309 - 20dynamics of cancer

294 CHAPTER 14

cell varies stochastically but, on average, the methylation frequency pro-
vides an indicator of the number of cellular divisions.
If bulge cells divide rarely and continue to be the ultimate progenitors
of hair renewal throughout life, then methylation will increase little with
age. In particular, the average methylation of follicles should rise very
early in life as cellular division during development creates the bulge
stem cells, then follicular methylation should remain nearly constant
during the remainder of life. Kim et al. (2006) found exactly that pattern:
increasing methylation up to around two years of age, followed by a long
plateau through the rest of life.
The bulge cells appear to be the ultimate stem cells in the follicle
hierarchy. If so, then in each hair cycle the bulge cells seed the follicular
germ with new daughter cells; those daughters act as stem cells for one
cycle and then die.
During each cycle, the follicular germ cells divide, and their daughter
transit lineages expand to produce the growing hair. The mitotic age of
cells temporarily rises as the hair cycle progresses.
Kim et al. (2006) analyzed whether mitotic age measurably increases
during a hair cycle by comparing methylation frequency between short
and long hairs. Short hairs tend to be earlier in a given hair cycle than
long hairs, and so the short hairs should on average have lower methy-
lation frequency. The observed methylation patterns match this predic-
tion of less methylation in short compared with long hairs. At the end
of the hair cycle, the follicular germ apparently dies off, to be reseeded
in the next cycle by relatively young and weakly methylated daughters
of the bulge cells.
These particular conclusions about mitotic age and stem cell hier-
archies remain tentative. The analysis does show clearly the potential
value of inferring lineage history from molecular markers.

VARIABLE LENGTH OF MICROSATELLITE REPEATS

Loss of DNA mismatch repair raises the mutation rate in repeated
DNA sequences. One type of repeat, the microsatellite, mutates often
in cells that are deficient in mismatch repair. I discuss two studies that
measured variation in microsatellite repeats among a set of cells at one
point in time, and used variation in those repeated regions to reconstruct
historical aspects of the cell lineages involved in tumorigenesis.