ANCESTRAL, MAMMALIAN-WIDE SUBFAMILIES OF LINE-1 REPETITIVE SEQUENCES

Analysis of the 3'-ends of approximately 900 separate human LINE-I (L1 ) elements from primates revealed 47 contiguous but distinct subfamili es with the LI family. Eight previously described medium reiteration f requency sequences (MERs) were found to be parts of ancient L1 untrans lated 3'-regions which show little or no sequence similarity to the pr esently active L1 3'-end. Some of the major changes in 3'-end sequence can be explained by recombination events between different L1 repeats as well as between L1 and unrelated repetitive sequences. One of thes e sequences, MER42, is reported in this paper. With the set of consens us sequences for different subfamilies and their diagnostic features, it is possible to estimate the age of individual LLNE-1 elements. Cont rary to earlier suggestions, the majority of L1 copies in the human ge nome is very old; more than half of the identifiable elements were ins erted into the genome before the mammalian radiation, as evidenced by elements at orthologous sites in human and other mammalian genomes. Mu ltiple distinct LI source genes seem to have been active simultaneousl y over long periods of time.