T. Ryan Gregory
Gene 324: 15-34.
Abstract
Numerous theories have been proposed to account
for the pronounced differences in the quantity of non-coding DNA among
eukaryotic genomes, but the current repertoire remains incomplete
because the only
explicit mechanisms it provides involve DNA gain. It has been
proposed
more recently that biases in spontaneous insertions and deletions
(indels)
can lead to genome shrinkage by mutational mechanisms alone. The
present
article provides the first detailed critical discussion of this
approach,
and covers three different ideas related to it: 1) the general notion
of
DNA loss by deletion bias, 2) the “DNA loss hypothesis” which supposes
that
variation in genome size can be attributed to differences in DNA loss
rate,
and 3) the “mutational equilibrium model” which attempts to describe
the
long-term evolution of genome size. The mutational equilibrium
model
is found to be problematic, and it is noted that DNA loss by small
indels
is too slow in real time to determine variation in genome size above a
relatively
low threshold. Some alternative explanations for the observed
patterns
are provided, and the critique also identifies some potential problems
with
the current dataset. These include a failure to cite a more
detailed
(and somewhat contradictory) mammalian dataset, a questionable use of
arithmetic means with highly skewed data, and important discrepancies
among the particular DNA sequences so far analyzed. Overall,
evolutionary reductions in
genome size are considered important, but the specific mechanism
relating
to small deletion bias is far too weak to be accepted as a primary
determinant
of genome size variation in general.