The introns-early view has been challenged for several genes; prominent ins
tances are triose phosphate isomerase (TPI), aldolase, pyruvate kinase (PK)
, alcohol dehydrogenase (ADH), glyceraldehyde-3-phosphate dehydrogenase (GA
PDH) and myosin heavy chain. While some of their introns appear to be phylo
genetically ancient and/or to delineate exons corresponding to protein modu
les, a considerable number seemingly do not. But it is argued here that man
y of these anomalous introns are periodic, that is, relies of internal sequ
ence repetitions within the ancestral gene. Some of these periodic-intron p
atterns are shared between related genes, as in the alpha beta -barrels of
TPI, aldolase and PK, or the Rossmann nucleotide-binding domain common to P
K, ADH and GAPDH. This is further evidence for the ancestral status of thes
e introns. The myosin heavy chain C-terminal rod region is paradoxical in t
hat its sequence is clearly periodic but its intron placements are not; how
ever, they exhibit a remarkable coherence of intron translational phases, s
uggesting that these introns may also have originally had a periodic arrang
ement now obscured by intron slipping. (C) 2000 Academic Press.