Species-specific homogeneity of the primate Alu family of repeated DNA sequences

We have determined the base sequence of several cloned Alu family members from the DNAs of a new world monkey (owl monkey) and a prosimian (galago). The three owl monkey Alu family members reported here belong to a single 300 base pair consensus sequence which closely resembles the human Alu family consensus. The galago Alu family members can best be represented as belonging to either of two related but distinct consensus sequences. One of the two galago Alu family subgroups (Type I) more accurately resembles the human consensus sequence than does the other subgroup (Type II). In this work we compare base sequences of human and galago Type I Alu family members. There are several examples of species-specific differences between the human and Type I galago sequences indicating that the Alu family members are effectively homogenized within a species.     

A human-specific subfamily of Alu sequences.

Of a total of 500,000 Alu family members, approximately 500 are present as a human-specific (HS) subfamily. Each of the HS subfamily members shares a high degree of nucleotide identity and is not present at orthologous positions in other primate genomes, suggesting that HS subfamily members have recently inserted within the human genome. This confirms the hypothesis that the majority of Alu family members are amplified copies of a "master" gene(s). This master gene appears to be amplifying at a rate much slower than that seen earlier in primate evolution. Some of the HS Alu subfamily members have amplified so recently that they are dimorphic in the human population, making them a potentially powerful tool for studies of human populations.     

A high frequency of mtDNA polymorphisms in HeLa cell sublines

The complete mtDNA sequences from the uncloned "founder" HeLa cells and from five sublines have been determined. These sequences all carry a common "core" of 38 single basepair alterations relative to the revised Cambridge Reference Sequence (CRS). The HeLa mitochondrial genome is of African descent and it is a member of the African L3 haplogroup. The sequence of the HeLa mtDNA resolves the uncertainty surrounding the mosaic composition of the original CRS for human mtDNA. Most importantly, we detected a total of eight polymorphisms that have arisen in the mtDNA coding region of different HeLa sublines. These observations suggest that HeLa mtDNA has a high rate of sequence divergence, relative to the phylogenetically-derived divergence rate for mtDNAs in the human population, which results from a relaxation of negative selection against the fixation of deleterious mutations. Furthermore, this high frequency of polymorphisms in HeLa mtDNA may reflect a process similar to the accumulation of somatic mtDNA mutations in human cancers. Preliminary analysis of single-cell derived subclone lines revealed the occurrence of another polymorphism and provided evidence for a large number of mtDNA segregation units.     

Interspersed repeated sequences in the African green monkey genome that are homologous to the human Alu family.

The dominant family of interspersed repetitive DNA sequences in the human genome has been termed the Alu family. We have found that more than 75% of the lambda phage in a recombinant library representing an African green monkey genome hybridize with a human Alu sequence under stringent conditions. A group of clones selected from the monkey library with probes other than the Alu sequence were analyzed for the presence and distribution of Alu family sequences. The analyses confirm the abundance of Alu sequences and demonstrate that more than one repeat unit is present in some phages. In the clones studied, the Alu units are separated by an average of 8 kilobase pairs of unrelated sequences. The nucleotide sequence of one monkey Alu sequence is reported and shown to resemble the human Alu sequences closely. Hence, the sequence, dispersion pattern, and copy number of the Alu family members are very similar in the African green monkey and human genomes. Among the clones investigated were two that contain segments of the satellite DNA term alpha-component joined to non alpha-component DNA. The experiments indicate that in the monkey genome Alu sequences can occur close to regions of alpha-component DNA.     

DNA sequence polymorphisms in Alu repeats

We have developed an efficient method for detection of sequence differences in genomic DNA based on a new principle (M. Orita et al., 1989, Genomics 5: 874-879). Using this method, we show here that approximately half the Alu repeats interspersed in the human genome are significantly polymorphic. Analysis of Alu repeat polymorphism should be useful in construction of a high-resolution map and also in identifying genotypes of individuals for clinical and other purposes because the repeats are ubiquitous and the technique for their detection is simple.     

Role of poly(A) tail length in Alu retrotransposition

Alu are mobile noncoding Short INterspersed Elements (SINEs) present at a million copies in the human genome. Using marked Alu sequences in an ex vivo assay, we previously showed that they are mobilized through diversion of the LINE (Long INterspersed Elements) retrotransposition machinery, with the poly(A) tail of the Alu being required for their mobility. Here we show that other homopolymeric tracts cannot functionally replace the Alu poly(A) tail, and that the Alu transposition rate varies over a two-log range depending on the poly(A) tail length. Variation is according to a sigmoid-shaped curve with a lag observed for tails shorter than 15 nt and a plateau reached for tails longer than 50 nt, consistent with the binding of a limited number of a protein component requiring multiple contacts for a productive interaction with the poly(A) stretch. This analysis indicates that most of the naturally occurring genomic Alu, owing to their pA tail length, should be poor substrates for the LINE machinery, a feature possibly "selected" for the host sake.     

The multiple origins of human Alu sequences

Summary I have analyzed a collection of published human Alu sequences. The compiled sequences show several unexpected features, including a uniform pattern of divergence from their consensus sequence, a mutual divergence that is not correlated with their age, and common features in the genomic DNA flanking the 5 ends of the elements. I suggest that the Alu family of sequences derives from a large pool of precursors and not from a single precursor similar to the family consensus sequence, and that new elements integrate into the genome selectively at oligo-A-rich sites.     

Human Alu insertion polymorphisms in North African populations

Several features make Alu insertions a powerful tool used in population genetic studies: the polymorphic nature of many Alu insertions, the stability of an Alu insertion event and, furthermore, the ancestral state of an Alu insertion is known to be the absence of the Alu element at a particular locus and the presence of an Alu insertion at the site that forward mutational change. This study analyses seven Alu insertion polymorphisms in a sample of 297 individuals from the autochthonous population of Tunisia (Thala, Smar, Zarzis, and Bou Salem) and Libya with the aim of studying their genetic structure with respect to the populations of North Africa, Western, Eastern and Central Europe. The comparative analyses carried out using the MDS and AMOVA methods reveal the existence of spatial heterogeneity, and identify four population groups. Study populations (Libya, Smar, Zarzis, and Bou Salem) are closest to North African populations whereas Thala is isolated and is closest to Western European populations. In conclusion, Results of the present study support the important role that migratory movements have played in the North African gene pool, at least since the Neolithic period.     

Restriction fragment length polymorphisms detected in N-ras-related sequences of rats and their linkage analyses

Novel restriction fragment length polymorphisms (RFLPs) in inbred rats were revealed with the human N-ras gene as probe. Three fragments hybridizing to the probe were detected by Southern blot hybridization under highly stringent conditions, and one of the fragments showed variation in inbred rat strains. Furthermore, on hybridization under low-stringency conditions, an additional fragment hybridizing to the probe was observed, and this fragment also showed interstrain variation. These two variant fragments showed different distributions in 27 inbred rat strains and segregated in backcross progeny as codominant alleles of independent single autosomal loci. Therefore, the loci for these RFLPs were named Nras-1 and Nras-2, respectively. Analyses of linkages between the RFLPs and 11 other loci revealed that the Nras-2 locus was closely linked to the c locus (3.7 +/- 2.6%), which belongs to rat linkage group I.     

A yeast centromere acts in cis to inhibit meiotic gene conversion of adjacent sequences

The centromere of chromosome III (CEN3) of yeast has been examined for its ability to inhibit meiotic recombination in adjacent sequences. The effect of the centromere was investigated when it was adjacent to both of the recombining sequences (homozygous) or adjacent to only one of the two recombining DNA segments (hemizygous). When homozygous, CEN3 exerts a bidirectional repression of crossing over and a strong inhibition of gene conversion. This suggests that CEN3 reduces the frequency of crossing over by interfering with the initiation of proximal recombination events. When hemizygous, CEN3 impairs the ability of adjacent sequences to act as the recipient of genetic information during gene conversion. These results support the idea that the initiating event in yeast meiotic recombination involves the recipient molecule.     

Structural genes adjacent to interspersed repetitive DNA sequences

The observation that repetitive and single copy sequences are interspersed in animal DNAs has suggested that repetitive sequences are adjacent to single copy structural gene sequences. To test this concept, single copy DNA sequences contiguous to interspersed repetitive sequences were prepared from sea urchin DNA by hydroxyapatite fractionation (repeat-contiguous DNA fraction). These single copy sequences included about one third of the total nonrepetitive sequence in the genome as determined by the amounts recovered during the hydroxyapatite fractionation and by reassociation kinetics. ³H-labeled mRNA from sea urchin gastrula was prepared by puromycin release from polysomes and used in DNA-driven hybridization reactions. The kinetics of mRNA hybridization reactions with excess whole DNA were carefully measured, and the rate of hybridization was found to be 3–5 times slower than the corresponding single copy DNA driver reassociation rate. The mRNA hybridized with excess repeat-contiguous DNA with similar kinetics relative to the driver DNA. At completion 80% of that mRNA hybridizable with whole DNA (approximately 65%) had reacted with the repeat-contiguous DNA fraction (50%). This result shows that 80–100% of the mRNA molecules present in sea urchin embryos are transcribed from single copy DNA sequences adjacent to interspersed repetitive sequences in the genome.     

Human satellite I sequences include a male specific 2.47 kb tandemly repeated unit containing one Alu family member per repeat

A portion of human satellite I DNA is digested by HinfI into three fragments of 775, 875 and 820bp in length which form a tandemly repeated unit 2.47kb in length, specific to male DNA. One Alu family member per repeat is found within the relatively G+C rich 775bp fragment. The 875 and 820bp fragments are highly A+T rich and consist of long stretches of poly dAdT and related sequences.     

Alu repeated DNAs are differentially methylated in primate germ cells.

A significant fraction of Alu repeats in human sperm DNA, previously found to be unmethylated, is nearly completely methylated in DNA from many somatic tissues. A similar fraction of unmethylated Alus is observed here in sperm DNA from rhesus monkey. However, Alus are almost completely methylated at the restriction sites tested in monkey follicular oocyte DNA. The Alu methylation patterns in mature male and female monkey germ cells are consistent with Alu methylation in human germ cell tumors. Alu sequences are hypomethylated in seminoma DNAs and more methylated in a human ovarian dysgerminoma. These results contrast with methylation patterns reported for germ cell single-copy, CpG island, satellite, and L1 sequences. The function of Alu repeats is not known, but differential methylation of Alu repeats in the male and female germ lines suggests that they may serve as markers for genomic imprinting or in maintaining differences in male and female meiosis.     

Small RNA molecules related to the Alu family of repetitive DNA sequences.

A rodent 4.5S RNA molecule with extensive homology to the Alu family of interspersed repetitive DNA sequences has been found physically associated with polyadenylated nuclear and cytoplasmic RNAs (W. Jelinek and L. Leinwand, Cell 15:205-214, 1978; S. Haynes et al., Mol. Cell. Biol. 1:573-583, 1981). In this report, we describe a 4.5S RNA molecule in rat cells whose RNase fingerprints are identical to those of the equivalent mouse molecule. We show that the rat 4.5S RNA is part of a small family of RNA molecules, all sharing sequence homology to the Alu family of DNA sequences. These RNAs are synthesized by RNA polymerase III and are developmentally regulated and short-lived in the cytoplasm. Of this family of small RNAs, only the 4.5S RNA is found associated with polyadenylated RNA.     

Specific binding of Alu sequences by HeLa nuclear extracts

Plasmid Blur 8 which contains the 300bp human Alu consensus sequence and plasmid pBR322 were digested with restriction enzymes and the fragments obtained end labelled with 32P-gamma-ATP. The end labelled fragments were incubated with HeLa nuclear extracts and the incubation mixtures passed through a nitrocellulose filter. The 300bp alu consensus sequence was preferentially retained on the filter. The HeLa nuclear extract did not preferentially bind any fragments generated from pBR322 and histones which bind nonspecifically all DNA fragments did not preferentially bind the alu sequence. We conclude that the HeLa nuclear extract contains components which specifically bind the human alu sequence.     

Restriction fragment length polymorphism caused by a deletion involving Alu sequences within the human alpha 2-plasmin inhibitor gene

A restriction fragment length polymorphism within the human alpha 2-plasmin inhibitor gene has been detected by Southern blot hybridization using an alpha 2-plasmin inhibitor cDNA probe. This restriction fragment length polymorphism can be attributed to the presence of two alleles, A and B, that are distributed in Hardy-Weinberg equilibrium with frequencies of 73.5% and 2.65%, respectively, in 66 unrelated Caucasian individuals or with frequencies of 51.0% and 49.0%, respectively, in 50 unrelated Japanese individuals. The minor allele, B, is due to a deletion of about 720 base pairs in intron 8 of the alpha 2-plasmin inhibitor gene. Sequence analysis of the deletion junction in allele B and the corresponding regions of allele A demonstrated the presence of oppositely oriented Alu sequences at the 5' and 3' deletion boundaries. These data suggest that this restriction fragment length polymorphism was caused by intrastrand recombination between Alu sequences.     

Preferential expression of unique sequences adjacent to middle repetitive sequences in mouse cytoplasmic RNA.

Total single-copy DNA and single-copy DNA contiguous to middle repetitive sequences were isolated from mouse brain by successive hydroxylapatite column chromatographies. These DNAs, termed repeat-contiguous single-copy DNA, were found to constitute 48% of the total single-copy DNA. The saturation hybridization values of these two DNA probes to nuclear RNA and cytoplasmic RNA containing polyA of mouse brain and liver were measured. The saturation hybridization levels of total single-copy DNA to brain and liver nuclear RNA were 13.5% and 8.8%, respectively, and those of repeat-contiguous single-copy DNA to the same RNA samples were 13.3% and 8.5%, respectively. On the contrary, the saturation hybridization levels of single-copy DNA to cytoplasmic RNA containing polyA of brain and liver were 3.8% and 2.0%, respectively, and those of repeat-contiguous single-copy DNA to the same RNA samples were 5.8% and 4.0%, respectively. Similar results were obtained with total cytoplasmic RNA. These results indicate that about half the steady state nuclear RNA is transcribed from repeat-contiguous single-copy DNA, and that cytoplasmic RNA containing polyA is mainly derived from repeat-contiguous single-copy DNA.