SCIENTIFIC CORRESPONDENCE

Once a critical number of repeat units 1•8 The birth of has arisen in a given , that can become hypervariable, with SIR - Microsatellites, very short (2-5 base­ A separate nucleotide substitution occurring even on a generational time­ pair) tandem nucleotide repeats, are found occurred within the New World monkeys scale5. A broader understanding of micro­ scattered throughout the of along the lineage leading to owl mon­ satellite birth and early will eukaryotes1.2. Those loci with a high num­ key. This A-to-G transition created a require the sequencing of regions surroun­ ber of repeat units tend to vary in repeat dinucleotide repeat, (GT)5 or (TG)5, ding other expanded loci and analysis of number among individuals in a species, which has expanded by one repeat unit in these data in proper phylogenetic context. making them useful in many types of mod­ this species. 'Y]-Globins from other New Walter Messier em genetic studies3• Microsatellites have World monkeys that are more closely Shou-Hsien Li gained notoriety through their association related to this owl monkey need to be Caro-Beth Stewart with certain genetic diseases in humans4-6• sequenced to trace the evolution of this Department of Biological Sciences, Much has been written about the probable new more completely, esp­ University at Albany, mechanisms causing length variation within ecially to look for additional expansions. State University of New York, species2•6•7, but little is known about the evo­ Even though a wide phylogenetic spread Albany, New York 12222, USA lutionary origins of microsatellites 1• Here of primates has three contiguous GTs and 1. Valdes, A. M ., Slatkin, M. & Frei mer, N. B. 133, we document the independent 'birth' of two four contiguous TGs in this region of the 737-748 (1993). 2. Charlesworth, B. et al. Nature 371, 215-220 (1994). separate microsatellites in a short region of 'Y]-globin , of the species 3. Dowling, T. E. et al. in Molecular Systematics (eds Hillis, the 'Y]-globin pseudogene in primates. sampled, only the owl monkey has a D. M. et al.) Ch. 8 (Sinauer, Sunderland, MA, 1996). 4. Burke, J. R. et al. Nature Med. 2, 347-350 (1996). As shown in the figure, in many species dinucleotide expansion. Thus, a minimum 5. Campuzano, V. et al. Science 271, 1423-1427 (1996). of primates the DNA sequence in this number of repeat units may be necessary 6. Kuhl, D. P. A. & Caskey, C. T. Curr. Opin. Genet. Dev. 3, 404-407 I 1993). region is a single nucleotide substitution before initial expansion occurs. In this 7. Rubinsztein, D. C. et al. Nature Genet. 10, 337-343 away from the creation of two different dinucleotide repeat case, the minimum (1995). types of simple repeat sequences. Within number appears to be four or five, and in 8. Armour, J. A. L. et al. Hum. mo/ec. Genet. 3, 599-605 (1994). the hominoids, a single nucleotide substi­ the tetranucleotide repeat case, two. The 9. Koop, B. F. et al. Nature 319, 234-238 (1986). tution of a G to an A occurred along the evolutionary coupling of the nucleotide lineage leading to the common ancestor substitutions and the expansions suggests of the African apes (gorilla, bonobo and that the point mutations created enough Cosmological chimpanzee) and humans. This substi­ repeat units for subsequent microsatellite tution created a tetranucleotide repeat, expansion. The expansions we have found constant (ATGTh, which is expanded to (ATGT)4 in this region of the 'Y]-globin are con­ in the African apes and to (ATGT)5 in sistent with a stepwise model for SIR - Taken in sum 1, an impressive vari­ humans. As only one sequence is available microsatellite evolution 1•2 in which short ety of observations indicate that the cos­ from each species, we do not know if arrays increase or decrease in repeat num­ mological constant (_/1) is not zero. intraspecific variation in repeat number ber by one or two repeats, probably by Principal among these is the existence of exists at this locus. However, such poly­ replication slippage. We propose that ini­ stars that are more than 15 billion years morphism is unlikely, because loci with tial expansion of short repeat sequences, old2, in a Universe which is apparently fewer than about 5 to 8 repeat units are such as we document here, occurs relative­ much younger (less than 12 billion years almost never polymorphic in humans1•8• ly slowly on an evolutionary timescale. old), based on recent determinations of the Hubble constant3. A cosmological con­ stant increases the age of the Universe, but Human AAATTG'1'ATGiifTGT!ATG'1'lATG'l'\ATG'l'GATGACTGGGGAC only modestly in a flat Universe. Even with Chimpanzee AAATTG'flATGT!ATGTjATGTIA.TG~----GATGACTGGGGAC. a cosmological constant of !2,1==0.7 (where Bonobo AAATTG$TGT~TGT~TGT\ATGn----GATGACTGGGGAC !2,1 is expressed as a fraction of the critical AAATTGTIATGT,TGT~TGT(ATGn----GATGACTGGGGAC density), concordance between these ages AAATTGTATGTGTGT------GATGACTGGGGAC relies heavily on measurement errors2•3• AAATTGTATGTGTGT------GATGACTGGGGAC Moreover, the cosmological constant AAATTGTATGTGTGT------GATGACTGGGGAC has a strong effect4 on the deceleration AAATTGTATGTGTGT------GATGACTGGGGAC parameter q0, which provides an indepen­ AAATTGTATGTGTGT------GATGACTGGGGAC dent test of concordance. In particular, AAATTGTATGTGTGT------GATGACTGGGGAC q0 == l/2!2m-!2,1, where Dm is the density of matter as a fraction of the critical density. AAATTGTATGTGTGT------GATAACTGGGGAC If nm == 0.3 and !2,1 == 0.7 as suggested1, ---.- Owl monkey AAAT'I'G'l'k3TIGTG'l1G'l1G'D------GATAACTGGGAAC then q0 == -0.55. Although presently A-> G fraught with uncertainty, attempts5 to mea­ A short portion of a longer alignment of all available nucleotide sequences of the 'T]-globin sure q0 directly usually suggest positive, pseudogene of primates. The multiple alignment of these primate sequences is unambiguous rather than negative, values. If q0 ~ 0 in a surrounding this region (which encompasses nucleotides 1,311-1,350 of the full-length human flat Universe, D,1 s; 1/2!2m, and the cosmo­ sequence, and would be in the second of an expressed globin gene). Gaps relative to the logical constant cannot be significant. human sequence are indicated by dashes, Those nucleotide substitutions (indicated by dots) that created repeat units are mapped onto the known phylogeny of these primates in the man• Stacy S. McGaugh ner that requires the fewest events, Boxes delineate microsatellite repeats. For the sake of Department of Terrestrial Magnetism, clean alignment to the human sequence, we have boxed the GT repeats in the owl monkey; Carnegie Institution of Washington, however, it is just as likely that this expansion involved a TG repeat, as mutations in the homi• Washington, DC 20015, USA noids should have no effect on those in the New World monkeys. GenBank accession numbers: 1. Ostriker, J.P. & Steinhardt, P. J. Nature 377, 600-602 Human, X02133; chimpanzee, K02542; bonobo, X79724; gorilla, K02543; orangutan, (1995). M18038; gibbon, M54985; owl monkey, X02142; spider monkey, J03050; white-fronted 2. Bolte, M. & Hogan, C. J. Nature 376, 399-402 (1995). capuchin, M81409. Rhesus macaque sequence is from ref. 9. Sequences from patas monkey, 3. Freedman, W. L. et al. Nature 371, 757-762 (1994). 4. Peebles, P. E. J. Principles of Physical Cosmology Angolan colobus and Hanuman langur were generated in our laboratory by P. Banerjee; analysis (Princeton Univ. Press, 1993). of these will be published elsewhere. 5. Sandage, A. R. in The Deep Universe, Saas-Fee Advanced Course 23 (Springer, New York, 1995). NATURE · VOL 381 · 6 JUNE 1996 483