Introduction

Membership in the retroid family (Fuetterer & Hohn, 1987) is defined by the shared amino acid sequence identity to the conserved motifs in the reverse transcriptase (RT) segment of the RNA-directed DNA polymerase. The retroid family is composed of retro viruses, two different classes of DNA viruses, two distinct types of retrotransposons, retroposons, group II introns and plasmids of cellular organelles, an orphan group, the retrons (Temin, 1989) of bacteria (Inouye et ai, 1989; Lampson, Inouye & Inouye, 1989; Lim & Maas, 1989) and the telomere elongation protein (EST-1) of yeast (Table 27.1) (Lundblad & Blackburn, 1990).

The variability of gene content and sequence similarity in extant retroid elements (Table 27.1) can be used to infer a model of the evolutionary history of the retroid family (Fig. 27.1). The phylogenetic tree topology, based on RT similarities, indicates an initial bifurcation leading to lineages I and II. Independently derived phylogenies for each set of homologous proteins in the capsid/ribonuclear protein/protease/ reverse transcriptase/ribonuclease H/histidine-cysteine motif/integrase (CA/NC/PR/RT/RH/H-C/IN) element and capsid/reverse transcriptase/ tether (CA/RT/T) element produce trees that are congruent with the RT tree, suggesting that the ancestral units evolved as linked genes (Doolittle et ai, 1989; McClure, 1992; McClure, unpublished data). The hypothetical ancestral unit of lineage I is further supported by the congruency of the phylogenetic tree for the H-C/IN segment, even though it is found in a position in the copia-likt retrotransposons different from its position in all other retroid elements. Subsequent acquisition or deletion of gene units may have occurred as indicated along the various branches leading to the tree tips (Fig. 27.1).