Subfamily structure and evolution of the human L1 family of repetitive sequences

Using bioinformatic and phylogenetic approaches to classify transposable elements and understand their complex evolutionary histories

In recent years, much attention has been paid to comparative genomic studies of transposable elements (TEs) and the ensuing problems of their identification, classification, and annotation. Different approaches and diverse automated pipelines are being used to catalogue and categorize mobile genetic elements in the ever-increasing number of prokaryotic and eukaryotic genomes, with little or no connectivity between different domains of life. Here, an overview of the current picture of TE classification and evolutionary relationships is presented, updating the diversity of TE types uncovered in sequenced genomes. A tripartite TE classification scheme is proposed to account for their replicative, integrative, and structural components, and the need to expand in vitro and in vivo studies of their structural and biological properties is emphasized. Bioinformatic studies have now become front and center of novel TE discovery, and experimental pursuits of these discoveries hold great promise for both basic and applied science.

De novo 13q deletions in two patients with mild anorectal malformations as part of VATER/VACTERL and VATER/VACTERL-like association and analysis of EFNB2 in patients with anorectal malformations

Anorectal malformations (ARMs) comprise a broad spectrum of conditions ranging from mild anal anomalies to complex cloacal malformations. In 40-50% of cases, ARM occurs within the context of defined genetic syndromes or complex multiple congenital anomalies, such as VATER/VACTERL (vertebral defects [V], ARMs [A], cardiac defects [C], tracheoesophageal fistula with or without esophageal atresia [TE], renal malformations [R], and limb defects [L]) association. Here, we report the identification of deletions at chromosome 13q using single nucleotide polymorphism-based array analysis in two patients with mild ARM as part of VATER/VACTERL and VATER/VACTERL-like associations. Both deletions overlap the previously defined critical region for ARM. Heterozygous Efnb2 murine knockout models presenting with mild ARM suggest EFNB2 as an excellent candidate gene in this region. Our patients showed a mild ARM phenotype, closely resembling that of the mouse. We performed a comprehensive mutation analysis of the EFNB2 gene in 331 patients with isolated ARM, or ARM as part of VATER/VACTERL or VATER/VACTERL-like associations. However, we did not identify any disease-causing mutations. Given the convincing argument for EFNB2 as a candidate gene for ARM, analyses of larger samples and screening of functionally relevant non-coding regions of EFNB2 are warranted. In conclusion, our report underlines the association of chromosome 13q deletions with ARM, suggesting that routine molecular diagnostic workup should include the search for these deletions. Despite the negative results of our mutation screening, we still consider EFNB2 an excellent candidate gene for contributing to the development of ARM in humans.

Epigenetic histone modifications of human transposable elements: genome defense versus exaptation

BACKGROUND

Transposition is disruptive in nature and, thus, it is imperative for host genomes to evolve mechanisms that suppress the activity of transposable elements (TEs). At the same time, transposition also provides diverse sequences that can be exapted by host genomes as functional elements. These notions form the basis of two competing hypotheses pertaining to the role of epigenetic modifications of TEs in eukaryotic genomes: the genome defense hypothesis and the exaptation hypothesis. To date, all available evidence points to the genome defense hypothesis as the best explanation for the biological role of TE epigenetic modifications.

RESULTS

We evaluated several predictions generated by the genome defense hypothesis versus the exaptation hypothesis using recently characterized epigenetic histone modification data for the human genome. To this end, we mapped chromatin immunoprecipitation sequence tags from 38 histone modifications, characterized in CD4+ T cells, to the human genome and calculated their enrichment and depletion in all families of human TEs. We found that several of these families are significantly enriched or depleted for various histone modifications, both active and repressive. The enrichment of human TE families with active histone modifications is consistent with the exaptation hypothesis and stands in contrast to previous analyses that have found mammalian TEs to be exclusively repressively modified. Comparisons between TE families revealed that older families carry more histone modifications than younger ones, another observation consistent with the exaptation hypothesis. However, data from within family analyses on the relative ages of epigenetically modified elements are consistent with both the genome defense and exaptation hypotheses. Finally, TEs located proximal to genes carry more histone modifications than the ones that are distal to genes, as may be expected if epigenetically modified TEs help to regulate the expression of nearby host genes.

CONCLUSIONS

With a few exceptions, most of our findings support the exaptation hypothesis for the role of TE epigenetic modifications when vetted against the genome defense hypothesis. The recruitment of epigenetic modifications may represent an additional mechanism by which TEs can contribute to the regulatory functions of their host genomes.

Molecular archeology of L1 insertions in the human genome

BACKGROUND

As the rough draft of the human genome sequence nears a finished product and other genome-sequencing projects accumulate sequence data exponentially, bioinformatics is emerging as an important tool for studies of transposon biology. In particular, L1 elements exhibit a variety of sequence structures after insertion into the human genome that are amenable to computational analysis. We carried out a detailed analysis of the anatomy and distribution of L1 elements in the human genome using a new computer program, TSDfinder, designed to identify transposon boundaries precisely.

RESULTS

Structural variants of L1 elements shared similar trends in the length and quality of their target site duplications (TSDs) and poly(A) tails. Furthermore, we found no correlation between the composition and genomic location of the pre-insertion locus and the resulting anatomy of the L1 insertion. We verified that L1 insertions with TSDs have the 5'-TTAAAA-3' cleavage site associated with L1 endonuclease activity. In addition, the second target DNA cut required for L1 insertion weakly matches the consensus pattern TTAAAA. On the other hand, the L1-internal breakpoints of deleted and inverted L1 elements do not resemble L1 endonuclease cleavage sites. Finally, the genome sequence data indicate that whereas singly inverted elements are common, doubly inverted elements are almost never found.

CONCLUSIONS

The sequence data give no indication that the creation of L1 structural variants depends on characteristics of the insertion locus. In addition, the formation of 5' truncated and 5' inverted L1s are probably not due to the action of the L1 endonuclease.

Tel-2 is a novel transcriptional repressor related to the Ets factor Tel/ETV-6

We report here the isolation of Tel-2, a novel member of the Ets transcription factor family, with high homology to Tel/ETV-6. Tel-2 is the second mammalian member of the Tel Ets family subclass whose prototype Tel is involved in various chromosomal translocations in human cancers. Six differentially expressed alternative splice products of Tel-2 were characterized encoding different Tel-2 isoforms which either contain or lack the amino-terminal Pointed domain and also vary at the carboxyl terminus. In contrast to Tel, which is highly expressed in several different cell types and tissues, Tel-2 is only weakly expressed in a variety of tissues and cell types, including placenta, prostate, spleen, liver, and lung. Tel-2 binds to functionally relevant Ets-binding sites of several genes and only the Tel-2 isoform containing the Pointed domain and the DNA-binding domain acts as a strong repressor of transcription. The retinoic acid receptor alpha and bone morphogenetic protein-6B (BMP-6) genes are specifically repressed by Tel-2 indicating a function for Tel-2 as an inhibitor of differentiation. Due to the important involvement of Tel in human cancer and the location of Tel-2 within the MHC cluster region, Tel-2 might be involved in chromosomal translocations in human cancer as well.

Complete sequence and polymorphism study of the human TYRP1 gene encoding tyrosinase-related protein 1

The complete 24,667 nucleotide sequence spanning the human TYRP1 gene has been determined from the inserts of two overlapping lambda clones. A LINE-1 repeat element is immediately adjacent to and may demarcate the immediate 5' promoter region of the gene. A search for polymorphism within the seven TYRP1 coding exons has been performed by an RNase mismatch detection procedure. Analysis of the TYRP1 gene in 100 Caucasian individuals of varying hair color has found no amino acid sequence variation nor revealed any hemizygous mutant allele in the hypopigmented phenotype of two 9p- syndrome patients.

Repetitive sequence fingerprinting in the long range mapping of mammalian genomes

This review presents some properties of interspersed repeats, particularly human and mouse repeats, and shows how these have been utilized in long-range genome mapping. The link between the distribution of such repeats and their relationship with genome organization is discussed.

A tandemly repeated DNA family originated from SINE-related elements in the European plethodontid salamanders (Amphibia, Urodela)

We have characterized a highly repetitive family, named Hy/Pol III, in the genome of the European salamanders Hydromantes (Plethodontidae). This family consists of short, tandemly repeated sequences organized in clusters, scattered through the genome as shown both by in situ hybridization to chromosomes and by Southern blot hybridization. The repeat unit is about 200 bp in length and it is a composite element since it contains a SINE-like retroposon with a tRNA structure, flanked by two short direct repeats. The whole element itself is bordered by two other direct repeats. The sequence data suggest that two elements, presumably derived from polymerase III transcripts, have been inserted one into the other, giving rise to the observed composite structure. During evolution the Hy/Pol III family was then amplified by tandem duplication at the DNA level. The inferred relationships between Hy/Pol III members from three representative species of the European Hydromantes suggests that a subfamily structure characterizes the evolutionary history of this family.

Master genes in mammalian repetitive DNA amplification

The analysis of species-specific subfamilies of both the LINE and SINE mammalian repetitive DNA families suggests that such subfamilies have arisen by amplification of an extremely small group of 'master' genes. In contrast to the master genes, the vast majority of both SINEs and LINEs appear to behave like psudogenes in their inability to undergo extensive amplification.

Fractals in biology and medicine

Our purpose is to describe some recent progress in applying fractal concepts to systems of relevance to biology and medicine. We review several biological systems characterized by fractal geometry, with a particular focus on the long-range power-law correlations found recently in DNA sequences containing noncoding material. Furthermore, we discuss the finding that the exponent alpha quantifying these long-range correlations ("fractal complexity") is smaller for coding than for noncoding sequences. We also discuss the application of fractal scaling analysis to the dynamics of heartbeat regulation, and report the recent finding that the normal heart is characterized by long-range "anticorrelations" which are absent in the diseased heart.

Integration of a vector containing a repetitive LINE-1 element in the human genome

Mammalian cells contain numerous nonallelic repeated sequences, such as multicopy genes, gene families, and repeated elements. One common feature of nonallelic repeated sequences is that they are homeologous (not perfectly identical). Our laboratory has been studying recombination between homeologous sequences by using LINE-1 (L1) elements as substrates. We showed previously that an exogenous L1 element could readily acquire endogenous L1 sequences by nonreciprocal homologous recombination. In the study presented here, we have investigated the propensity of exogenous L1 elements to be involved in a reciprocal process, namely, crossing-overs. This would result in the integration of the exogenous L1 element into an endogenous L1 element. Of over 400 distinct integration events analyzed, only 2% involved homologous recombination between exogenous and endogenous L1 elements. These homologous recombination events were imprecise, with the integrated vector being flanked by one homologous and one illegitimate junction. This type of structure is not consistent with classical crossing-overs that would result in two homologous junctions but rather is consistent with one-sided homologous recombination followed by illegitimate integration. Contrary to what has been found for reciprocal homologous integration, the degree of homology between the exogenous and endogenous L1 elements did not seem to play an important role in the choice of recombination partners. These results suggest that although exogenous and endogenous L1 elements are capable of homologous recombination, this seldom leads to crossing-overs. This observation could have implications for the stability of mammalian genomes.

Integration of a vector containing a repetitive LINE-1 element in the human genome

Mammalian cells contain numerous nonallelic repeated sequences, such as multicopy genes, gene families, and repeated elements. One common feature of nonallelic repeated sequences is that they are homeologous (not perfectly identical). Our laboratory has been studying recombination between homeologous sequences by using LINE-1 (L1) elements as substrates. We showed previously that an exogenous L1 element could readily acquire endogenous L1 sequences by nonreciprocal homologous recombination. In the study presented here, we have investigated the propensity of exogenous L1 elements to be involved in a reciprocal process, namely, crossing-overs. This would result in the integration of the exogenous L1 element into an endogenous L1 element. Of over 400 distinct integration events analyzed, only 2% involved homologous recombination between exogenous and endogenous L1 elements. These homologous recombination events were imprecise, with the integrated vector being flanked by one homologous and one illegitimate junction. This type of structure is not consistent with classical crossing-overs that would result in two homologous junctions but rather is consistent with one-sided homologous recombination followed by illegitimate integration. Contrary to what has been found for reciprocal homologous integration, the degree of homology between the exogenous and endogenous L1 elements did not seem to play an important role in the choice of recombination partners. These results suggest that although exogenous and endogenous L1 elements are capable of homologous recombination, this seldom leads to crossing-overs. This observation could have implications for the stability of mammalian genomes.

Cloning of the region between HLA-DMB and LMP2 in the human major histocompatibility complex

The human MHC is one of the most extensively mapped regions of the human genome. Almost all of the class II region of the MHC has already been cloned in cosmids but a gap remained between the DMB and LMP2 genes. Previously, screening of several complete cosmid libraries had failed to bridge this gap, which may contain novel antigen processing or presentation genes. We constructed cosmid libraries from two different sources in order to clone the region: (a) a library with fourfold coverage made from flow-sorted human chromosome 6 DNA and (b) a library derived from a yeast artificial chromosome clone spanning the region. Using this saturation approach, cosmid clones were eventually isolated over the region of interest. A single bacteriophage P1 clone was also obtained spanning the region. The YAC, cosmid, and P1 physical maps were consistent and the distance between the DMB and LMP2 genes was measured as 70 kb. It is not clear why DMB to LMP2 is infrequently represented in cosmid libraries, but the clones that we have obtained will now enable us to search for new coding sequences.

Statistical mechanics in biology: how ubiquitous are long-range correlations?

The purpose of this opening talk is to describe examples of recent progress in applying statistical mechanics to biological systems. We first briefly review several biological systems, and then focus on the fractal features characterized by the long-range correlations found recently in DNA sequences containing non-coding material. We discuss the evidence supporting the finding that for sequences containing only coding regions, there are no long-range correlations. We also discuss the recent finding that the exponent alpha characterizing the long-range correlations increases with evolution, and we discuss two related models, the insertion model and the insertion-deletion model, that may account for the presence of long-range correlations. Finally, we summarize the analysis of long-term data on human heartbeats (up to 10(4) heart beats) that supports the possibility that the successive increments in the cardiac beat-to-beat intervals of healthy subjects display scale-invariant, long-range "anti-correlations" (a tendency to beat faster is balanced by a tendency to beat slower later on). In contrast, for a group of subjects with severe heart disease, long-range correlations vanish. This finding suggests that the classical theory of homeostasis, according to which stable physiological processes seek to maintain "constancy," should be extended to account for this type of dynamical, far from equilibrium, behavior.

Emergence of master sequences in families of retroposons derived from 7sl RNA

The past few years have brought new insight into the evolution of families of retroposons. These are composed of a very small number of master sequences able to duplicate, and a large majority of copies that are inactive for retroposition. During the course of time, successive replacements of master sequences have produced waves of amplification that are recognizable as subfamilies. In the Alu and the B1 families, one can distinguish two evolutionary periods. The first involves only monomeric elements that are now extinguished (fossil elements) and is characterized by deep remodeling of the sequences. This period ends, in primates, with the fusion of a free left and a free right Alu monomer, producing the first modern Alu dimeric element; in rodents it ends with a tandem duplication of 29 bp to create the first modern B1 element. The second period is characterized by a great stability of the master sequences. The observed turn-over of master sequences is still an enigma. However, analysis of the contemporary master sequences and of the oldest master sequences provide some clues. Here, we review the very first stages of the appearance of the Alu and the B1 families in mammalian genomes.

The human episome HALF1: structure of its genomic counterpart

A human episomal sequence (HALF1) has been identified by its ability to restore expression of hepatic functions when used to transfect a rat dedifferentiated cell line. The genomic equivalent of this human episome (gHALF1) and its flanking sequences were analyzed. HALF1 itself does not present the characteristics of a transposable element but half of its sequence corresponds to retroposons, including Alu and L1 repeats and a processed pseudogene, known to transpose via RNA intermediates. The structural characteristics of these different kinds of retroposons and their origin and evolution were analyzed.

Multiple L1 progenitors in prosimian primates: phylogenetic evidence from ORF1 sequences

One of the uncertainties regarding the evolution of L1 elements is whether there are numerous progenitor genes. We present phylogenetic evidence from ORF1 sequences of slow loris (Nycticebus coucang) and galago (Galago crassicaudatus) that there were at least two distinct progenitors, active at the same time, in the ancestor of this family of prosimian primates. A maximum parsimony analysis that included representative L1s from human, rabbit, and rodents, along with the prosimian sequences, revealed that one of the galago L1s (Gc11) grouped very strongly with the slow loris sequences. The remaining galago elements formed their own unique and strongly supported clade. An analysis of replacement and silent site changes for each link of the most parsimonious tree indicated that during the descent of the Gc11 sequence approximately two times more synonymous than nonsynonymous substitutions had occurred, implying that the Gc11 founder was functional for some time after the split of galago and slow loris. Strong purifying selection was also evident on the galago branch of the tree. These data indicate that there were two distinct and contemporaneous L1 progenitors in the lorisoid ancestor, evolving under purifying selection, that were retained as functional L1s in the galago lineage (and presumably also in the slow loris). The prosimian ORF1 sequences could be further subdivided into subfamilies. ORF1 sequences from both the galago and slow loris have a premature termination codon near the 3' end, not shared by the other mammalian sequences, that shortens the open reading frame by 288 bp. An analysis of synonymous and nonsynonymous substitutions for the 5' and 3' portions, that included intra- and inter-subfamily comparisons, as well as comparisons among the other mammalian sequences, suggested that this premature stop codon is a prosimian acquisition that has rendered the 3' portion of ORF1 in these primates noncoding.

The evolution of coexisting highly divergent LINE-1 subfamilies within the rodent genus Peromyscus

Two distinct members of the LINE-1 (L1) family in Peromyscus were characterized. The two clones, denoted L1Pm55 and L1Pm62, were 1.5 kb and 1.8 kb in length, respectively, and align to the identical region of the L1 sequence of Mus domesticus. Sequence similarity was on the order of 70% between L1Pm55 and L1Pm62, which approximates that between either Peromyscus sequence and Mus L1. L1Pm62 represents a more prevalent subfamily than L1Pm55. L1Pm62 exists in about 500 copies per haploid genome, while L1Pm55 exists in about 100 copies. The existence of major and minor subpopulations of L1 within Peromyscus is in contrast to murine rodents and higher primates, where L1 copy number is on the order of 20,000 to 100,000, and where levels of intraspecific divergence among L1 elements are typically less than 15-20%. Additional Peromyscus clones are similarly divergent from both L1Pm62 and L1Pm55, implying the existence of more than two distinct L1 subfamilies. The highly divergent L1 subfamilies in Peromyscus apparently have been evolving independently for more than 25 million years, preceding the divergence of cricetine and murine rodents. Investigations of the evolution of L1 within Peromyscus by restriction and Southern analysis was performed using species groups represented by the partially interfertile species pairs P. maniculatus-P. polionotus, P. leucopus-P. gossypinus, and P. truei-P. difficilis of the nominate subgenus and P. californicus of the Haplomylomys subgenus. Changes in L1 and species group taxonomic boundaries frequently coincided. The implications for phylogeny are discussed.

DNA sequence analysis of 66 kb of the human MHC class II region encoding a cluster of genes for antigen processing

The genomic sequence of a 66,109 bp long region within the human MHC has been determined by manual and automated DNA sequencing. From cDNA mapping and sequencing data it is known that this region contains a cluster of at least four genes that are believed to be involved in antigen processing. Here, we describe the genomic organization of these genes, which comprise two proteasome-related genes (LMP2 and LMP7), thought to be involved in the proteolytic degradation of cytoplasmic antigens and two ABC transporter genes (TAP1 and TAP2), thought to be involved in pumping of the degraded peptides across the endoplasmic reticulum membrane. Analysis of the sequence homology and the intron/exon structures of the corresponding genes suggests that one gene pair arose by duplication from the other. Comparison of the available sequence data from other organisms shows striking conservation (70 to 84%) of this gene cluster in human, mouse and rat. The presence of several potential interferon stimulated response elements (ISREs) is in agreement with the experimentally observed up-regulation of these genes with gamma-interferon.

Prototypic sequences for human repetitive DNA

We report a collection of 53 prototypic sequences representing known families of repetitive elements from the human genome. The prototypic sequences are either consensus sequences or selected examples of repetitive sequences. The collection includes: prototypes for high and medium reiteration frequency interspersed repeats, long terminal repeats of endogenous retroviruses, alphoid repeats, telomere-associated repeats, and some miscellaneous repeats. The collection is annotated and available electronically.

Origin of the Alu family: a family of Alu-like monomers gave birth to the left and the right arms of the Alu elements

The Alu dimeric elements are a common feature of the primate genomes, where they constitute a family of related sequences (1). The identification of a free left Alu monomer (FLAM) family plus a free right Alu monomer (FRAM) family suggests that the dimeric structure results from the fusion of a FLAM sequence with a FRAM sequence (2). Here, we describe a very old Alu-like monomeric family, referred to as FAM for fossil Alu monomer. This family arose from a 7SL RNA sequence and gave birth to the FLAM and FRAM families. From the results obtained, the evolution of the Alu family can be subdivided into two phases. The first phase, which involves only monomeric elements, is characterized by deep remodelling of the progenitor sequences and ends with the appearance of the first Alu dimeric element through the fusion of a FLAM and a FRAM element. The second phase, still in progress, starts with the first Alu dimeric element. This phase is characterized by the stabilization of the progenitor sequences.

The beta globin gene cluster of the prosimian primate Galago crassicaudatus: nucleotide sequence determination of the 41-kb cluster and comparative sequence analyses

The nucleotide sequence of the beta globin gene cluster of the prosimian Galago crassicaudatus has been determined. A total sequence spanning 41,101 bp contains and links together previously published sequences of the five galago beta-like globin genes (5'-epsilon-gamma-psi eta-delta-beta-3'). A computer-aided search for middle interspersed repetitive sequences identified 10 LINE (L1) elements, including a 5' truncated repeat that is orthologous to the full-length L1 element found in the human epsilon-gamma intergenic region. SINE elements that were identified included one Alu type I repeat, four Alu type II repeats, and two methionine tRNA-derived Monomer (type III) elements. Alu type II and Monomer sequences are unique to the galago genome. Structural analyses of the cluster sequence reveals that it is relatively A+T rich (about 62%) and regions with high G+C content are associated primarily with globin coding regions. Comparative analyses with the beta globin cluster sequences of human, rabbit, and mouse reveal extensive sequence homologies in their genic regions, but only human, galago, and rabbit sequences share extensive intergenic sequence homologies. Divergence analyses of aligned intergenic and flanking sequences from orthologous human, galago, and rabbit sequences show a gradation in the rate of nucleotide sequence evolution along the cluster where sequences 5' of the epsilon globin gene region show the least sequence divergence and sequences just 5' of the beta globin gene region show the greatest sequence divergence.

Duplication of the gamma-globin gene mediated by L1 long interspersed repetitive elements in an early ancestor of simian primates

Regions surrounding the single gamma-globin gene of galago and the duplicated gamma 1- and gamma 2-globin genes of gibbon, rhesus monkey, and spider monkey were sequenced and aligned with those from humans. Contrary to previous studies, spider monkey was found to have not one but two gamma-globin genes, only one of which (gamma 2) is functional. The reconstructed evolutionary history of the gamma-globin genes and their flanking sequences traces their origin to a tandem duplication of a DNA segment approximately 5.5 kilobases long that occurred before catarrhine primates (humans, apes, and Old World monkeys) diverged from platyrrhines (New World monkeys), much earlier than previously thought. This reconstructed molecular history also reveals that the duplication resulted from an unequal homologous crossover between two related L1 long interspersed repetitive elements, one upstream and one downstream of the single ancestral gamma-globin gene. Perhaps facilitated by the redundancy resulting from the duplication, the gamma-globin genes escaped the selective constraints of embryonically functioning genes and evolved into fetally functioning genes. This view is supported by the finding that a burst of nonsynonymous substitutions occurred in the gamma-globin genes while they became restructured for fetal expression in the common ancestor of platyrrhines and catarrhines.

Evolution of mouse B1 repeats: 7SL RNA folding pattern conserved

In a recent report mouse B1 genomic repeats were divided into six families representing different waves of fixation of B1 variants, consistent with the retroposition model of human Alu elements. These data are used to examine the distribution of nucleotide substitutions in individual genomic repeats with respect to family consensus sequences and to compare the minimal energy structures of the corresponding B1 RNAs. By an enzymatic approach the predicted structure of B1 RNAs is experimentally confirmed using as a model sequence an RNA of a young B1 family member transcribed in vitro by T7 RNA polymerase. B1 RNA preserves folding domains of the Alu fragment of 7SL RNA, its progenitor molecule. Our results reveal similarities among 7SL-like retroposons, human Alu, and rodent B1 repeats, and relate the evolutionary conservation of B1 family consensus sequences to selection at the RNA level.

Reconstruction and analysis of human Alu genes

The existing classification of human Alu sequences is revised and expanded using a novel methodology and a larger set of sequence data. Our study confirms that there are two major Alu subfamilies, Alu-J and Alu-S. The Alu-S subfamily consists of at least five distinct subfamilies referred to as Alu-Sx, Alu-Sq, Alu-Sp, Alu-Sc, and Alu-Sb. The Alu-Sp and Alu-Sq subfamilies have been revealed by this study. Alu subfamilies differ from one another in a number of positions called diagnostic. In this paper the diagnostic positions are defined in quantitative terms and are used to evaluate statistical significance of the observed subfamilies. Each Alu subfamily most likely represents pseudogenes retroposed from evolving functional source Alu genes. Evidence presented in this paper indicates that Alu-Sp and Alu-Sc pseudogenes were retroposed from different source genes, during overlapping periods of time, and at different rates. Our analysis also indicates that the previously identified Alu-type transcript BC200 comes from an active Alu gene that might have existed even before the origin of dimeric Alu sequences. The source genes for Alu pseudogene families are reconstructed. It is assumed that diagnostic differences between reconstructed source genes reflect mutations that have occurred in true source Alu genes under natural selection. Some of these mutations are compensatory and are used to reconstruct a common secondary structure of Alu RNAs transcribed from the source genes. The biological function of Alu RNA is discussed in the context of its homology to the elongation-arresting domain of 7SL RNA.(ABSTRACT TRUNCATED AT 250 WORDS)