(I'll follow Will's precedent in posting a bias alert, since I'm personally
involved in all of this)
Will Fischer has a valid point about different types of introns. Arguing that
spliceosomal introns are ancient by invoking the widely scattered bacterial
distribution of group I and other non-spliceosomal introns is a bit like
proposing that the metazoan ancestor had feathered wings because bats and flies
have wings too, just like birds. Of course, we are all aware that the radical
differences in these functionally analogous "wings" belie common ancestry, but
(understandably) many molecular biologists are not yet aware that the same is
true of most of the different "intron" types (e.g., see Lambowitz and Belfort,
1993, Ann. Rev. Biochem. 62: 587-622). Actually, there may be an evolutionary
relationship between the group II introns and the spliceosomal introns (see
Cavalier-Smith, 1991, cited by Will), but I don't see how this would be
rationalized with an ancient origin of spliceosomal introns.
>The controversy continues (see Hurst's News&Views in the Sept_29 >Nature), but
parsimony is on the side of the introns-late advocates.
Hurst's proposal of a causal relationship between outcrossing and intron density
deserves attention (its a theory-based proposal that is clearly testable! what a
concept!), but I would like to supplement Hurst's comments by noting some
important results that were not mentioned in his meeting review.
The case for an ancient origin of introns presented in Gilbert's lecture (at the
CIAR meeting in Halifax, August, 1994) was based entirely on his analysis of a
hypothetical 11-intron ancestral TPI gene. This 11-intron (12-exon) ancestor
was hypothesized on the basis of the 14 presently known TPI intron positions, by
assuming that 11 of these known intron positions represent ancient positions,
with the 3 remaining ones having "slid" a few codons from nearby "ancient"
positions. Gilbert claimed that this 11-intron gene structure exhibited
an important correspondence with the structure of the encoded protein.
The 11-intron TPI scenario continued to be defensible for only about 2 hours
after it was presented. Then, Jeff Palmer (from the Logsdon laboratory, Indiana
Univ.) revealed that 7 novel TPI intron positions had been discovered, raising
the current total from 14 to 21. Palmer also noted that Andrew Roger (Dalhousie
University) has recently sequenced two completely intronless protist TPI genes
that are outgroups to the other characterized eukaryotic TPIs. Palmer also
relayed recent results from the Walker lab (Queens Univ., Ontario) to the effect
that the TPI intron they found in the mosquito _Culex tarsalis_ two years ago
(Tittiger, et al., Nature 361, 470)-- interpreted enthusiastically by
introns-early advocates as an important ancient intron-- is actually very
restricted in its phylogenetic distribution, being found only within a subgroup
Thus, at the Halifax meeting it became clear that the 11-intron hypothetical
ancestral TPI gene is a fiction based on incomplete data and optimistic
suppositions about the antiquity of various intron positions. This scenario, and
the conclusions based on it, are no longer relevant, a point that should be kept
in mind while reading the otherwise valuable commentary by Hurst.