strands of the repeat region (Fig. 1). In-
deed, Volpe et al. report in this issue thattranscripts are produced from both strands
of the dh repeat, and these transcripts mark-edly increase in abundance in S. pombemutants of dcr1, the Dicer homolog, and
ago1, the Argonaute homolog (7 ). Most ofour centromeric small RNAs cluster within
or near these transcripts, suggesting thatRNAs produced from each strand of therepeat anneal to form dsRNA that is
yeast. To investigate this possibility, we
cleaved by Dicer into the small RNAs (7 ).
croRNAs (miRNAs) are two types of ϳ22–
Mutations in dcr1 and ago1 in S. pombe
tially growing S. pombe using a method
important roles as regulators of gene ex-
pression in eukaryotes (1). siRNAs derive
of Dicer cleavage products, i.e., ϳ22-nt
function (7 ). Accordingly, we refer to these
from the successive cleavage of long dou-
groups (4 ). Of 61 sequenced clones, 49
and suggest that they specify the epigenetic
targets during RNA interference in animals
ground in such cloning efforts. Each of the
remaining 12 sequences matched the S.pombe centromeric repeats (Fig. 1). Neurospora. miRNAs are processed from
regulate expression levels of cell fate de-
endogenous hairpin transcripts such that a
are from the dh repeat, an element that can
confer heterochromatic silencing on anoth-
each hairpin molecule. Certain Caenorhab-
er locus and is sufficient for centromere
in unicellular yeast (1, 3). EndogenoussiRNAs that could cleave complementarymRNAs are also described in plants andanimals. However, other cloned RNAs donot fall into these two classes and might beheterochromatic siRNAs. In fact, epigenet-ic modifications have been correlated withsmall RNAs in multicellular eukaryotes,such as methylation of promoter DNA dur-ing transcriptional silencing of Arabidopsistransgenes (8). DNA methylation, which isdownstream of H3 K9 methylation in Neu-rospora (9), might be a consequence of
Fig. 1. Heterochromatic siRNAs (sequences A through L) from S. pombe. The loci and
orientation of matches to representative repeats are indicated below each centromere
methylation. Therefore, sequence-specific
fragment. Sequence K only matches ChrIII, and sequences D, E, and L match other repeats on
targeting of histone modifications could be
ChrIII. Centromeric repeats (10) are in green (dg) and orange (dh). Although the innermost
centromeric repeats contain tRNA sequences, all tRNA fragments that were cloned map
ditis elegans miRNAs are known to direct
function along with the centromeric central
References and Notes
translational repression of mRNA targets need-
core (5, 6 ). None of the RNAs match other
1. G. Hutva´gner, P. D. Zamore, Curr. Opin. Genet. Dev.
ed for proper larval development, and numer-
heterochromatic regions, such as the cen-
12, 225 (2002).
2. E. C. Lai, Nature Genet. 30, 363 (2002).
ous plant and animal miRNAs are thought to
3. M. W. Rhoades et al., Cell, 110, 513 (2002).
play similar roles in other contexts by target-
sequences, or the mating type locus region
4. N. C. Lau, L. P. Lim, E. G. Weinstein, D. P. Bartel,
ing mRNAs for translation attenuation or de-
homologous to the dh repeat, although our
Science 294, 858 (2001).
5. M. Baum, V. K. Ngan, L. Clarke, Mol. Biol. Cell 5, 747
struction (1–3). The ribonuclease III protein
Dicer is required for the processing of both
plete. Because S. pombe centromeres are
6. N. Ayoub, I. Goldshmidt, R. Lyakhovetsky, A. Cohen,
large regions (40 to 100 kb) with homolo-
Genetics 156, 983 (2000).
7. T. A. Volpe et al., Science 297, XXX (2002).
8. M. F. Mette, W. Aufsatz, J. van der Winden, M. A.
domain) proteins, whose biochemical func-
Matzke, A. J. Matzke, EMBO J. 19, 5194 (2000).
tions are unclear, are also necessary for pro-
domain of one centromere or from multiple
9. H. Tamaru, E. U. Selker, Nature 414, 277 (2001).
10. Y. Nakaseko, Y. Adachi, S. Funahashi, O. Niwa, M.
Yanagida, EMBO J. 5, 1011 (1986).
11. We thank T. Volpe et al. for sharing results before
in that transcription of adjacent genomic
found in the genome of Schizosaccharomy-
Whitehead Institute for Biomedical Research and
Department of Biology, Massachusetts Institute of
Technology, Cambridge, MA 02142, USA. E-mail:
www.sciencemag.org SCIENCE VOL 297 13 SEPTEMBER 2002
In , million people died. Violence killed fewer than one million ofthem; famine contributed to about six million deaths; more than million died of some form of disease.1 Many of these illnesses were theresult primarily of old age and may have been unpreventable. That isunlikely to be true, however, of over a million deaths from malaria, nearlytwo million deaths from tuberc
(Saguraku- enena--- sinonim rijeÄ i saguraku- koji se kod Zeamija joÅ¡ uvijek oznaÄ ava kao Noumjetnost. RijeÄ enen- doslovno, produžetak života, nekad je oznaÄ avala skup plesovaizvoÄ enih u budistiÄ kim hramovima povodom praznika, a saguraku je tkÄ . bio dio programa. Tokom godina stekla uže znaÄ enje- obredni ples pojedinih hramova)Kad se veliko božanstvo povuklo u svoj k