Sequence conservation in

A statistical analysis of a set of genomic human Alu elements is based on a published alignment and a recent classification of these sequences. After separation of the Alu sequences into families, the consensus sequences of these families are determined, using the correct weighting of the unidirectional decay of CG-dinucleotides. For, the tenfold greater mutation rate at CG's requires separate consideration of an independent clock at every stage of analysis. The distributions of the substitutions with respect to the new consensus sequences, taking the CG and the non-CG-nucleotide positions separately, lie far closer to the expected distributions than the total diversity. Computer analysis of the folding of RNAs derived from these sequences indicates that RNA secondary structure is conserved among Alu families, suggesting its importance for Alu proliferation and/or function. The folding pattern, further substantiated by a number of compensatory mutations, includes secondary structure domains which are homologous to those observed in 7SL RNA and a defined region of interaction between the two Alu subunits. These results are consistent with a model in which a small number of conserved Alu master genes give rise via retroposition to the numerous copies of Alu pseudogenes, that then diversify by random substitution. The master genes appeared at different periods during evolution giving rise to different families of Alu sequences. INTRODUCTION More than half a million AJu elements (1,2) are interspersed throughout the human genome by retroposition, i.e. by reintegration of reverse-transcribed copies of Alu RNAs (3). Alu elements are composed of two subunits, left and right, both homologous to the 7SL RNA sequence truncated by internal deletions (4). Although no function has been ascribed to this genetic material, its high genomic representation and recent reports of Alu involvement in mutations causing genetic defects (5-8) indicate an important role in primate evolution. Despite variability, all Alu elements can be related to an average consensus sequence, as done by Kariya et al. (9) from an alignment of 50 Alu sequences (for a review see ref.3). Recently, Jurka and Smith (10; cf. also 11 -14) were able to classify Alu elements into families, J, Sa, Sc and Sb, based on the presence of correlated nucleotide substitutions in a number of sequence positions. The age of these families is in the above order, as judged both by their mutational diversity and their divergence from the 7SL RNA sequence. Inspection of the Alu alignment indicates a high proportion of TG and CA dinucleotides in sequence positions corresponding to a CG dinucleotide in the consensus (cf. also 11). This can be explained by a higher transition rate of C to T due to CG methylation at the DNA level (15). A quantitative analysis of this CG-effect on Alu elements seems warranted, for, if this effect is of the order of 10:1 in pseudogenes (16), it should be taken into account in the compilation of their consensus sequences and when analyzing their sequence diversity.