© 2003 Nature Publishing Group http://www.nature.com/naturegenetics
B R I E F C O M M U N I C AT I O N S
Imprinted microRNA genes transcribed antisense to a reciprocally imprinted retrotransposon-like gene Hervé Seitz1, Neil Youngson2, Shau-Ping Lin2, Simone Dalbert3, Martina Paulsen3, Jean-Pierre Bachellerie1, Anne C Ferguson-Smith2 & Jérôme Cavaillé1 MicroRNAs (miRNAs) are an abundant class of RNAs that are ∼21–25 nucleotides (nt) long, interact with mRNAs and trigger either translation repression or RNA cleavage (RNA interference, RNAi) depending on the degree of complementarity with their targets. Here we show that the imprinted mouse distal chromosome 12 locus encodes two miRNA genes expressed from the maternally inherited chromosome and antisense to a retrotransposon-like gene (Rtl1) expressed only from the paternal allele. We mapped four recently reported mouse tissue-specific miRNA genes1 to an imprinted locus on mouse distal chromosome 12/human 14q32 (Fig. 1). Two of these, mir136 and mir-127, are located near two CpG islands in a mouse retrotransposon-like gene, Rtl1, which is expressed exclusively from the paternal allele (Fig. 2a,b). These miRNA genes are transcribed in an antisense orientation to Rtl1 and are expressed exclusively from the maternal chromosome (Fig. 2a,b). The mechanism by which miR-127 and miR136 are generated has yet to be determined, but miR-127 does not seem to be processed from long double-stranded RNAs formed by Rtl1 and anti-Rtl1 RNAs, as we did not detect any RNA species ∼21–23 nt long corresponding to anti-miR-127 (data not shown). Moreover, RT–PCR analysis indicates that these miRNAs are generated from a larger host transcript (Fig. 2b). The other two miRNA genes, mir-154 and mir-134, are located in a cluster of conserved repeated sequences2 (A-repeats; Fig. 1). The gene mir154 (A13) and several copies of related
sequences are intron-encoded in Mirg (microRNA-containing gene), a novel maternally expressed gene that lacks an open reading frame (Figs. 1 and 2c). As we could not detect them by northern-blot analysis, their imprinted status and mode of biosynthesis from Mirg introns have yet to be resolved. Although some Rtl1 RNA is clearly detectable (Fig. 2a), the perfect complementarity between miR-136, miR-127 and the Rtl1 open reading frame (Fig. 2d) suggests they might function as small interfering RNAs3,4 to silence Rtl1, presumably in a tissue-specific manner. For instance, such a role could be achieved in the embryo brain or in the placenta, in which mir-127 and Rtl1 are co-expressed and reciprocally imprinted (data not shown). Moreover, an involvement of miRNAs in Rtl1 imprinting cannot be formally ruled out. Although RNA interference is believed to be a trans-acting mechanism, preferential nuclear degradation of the Rtl1 transcript from the maternal allele only, mediated by the sense/antisense gene organization of the Rtl1 miRNA genes, could account for Rtl1 imprinting. Alternatively or additionally, miRNAs might mediate chromatin epigenetic changes specifically at the maternal Rtl1 allele, consistent with the involvement of the RNAi machinery in heterochromatin formation5–7. RNA-directed allele-specific methylation of the two CpG islands in Rtl1 is unlikely, as they are hypermethylated on both alleles and this occurs even in pUPD12 embryos in which miRNAs
C/D snoRNA gene miRNA gene cluster Dlk1 Meg3
cluster
Rtl1
(A-repeats)
Dio3
Mouse distal 12
Pat Mat mir-127
mir-136 mir-134 2 1
4 3
6 5
8 7
10 9
mir-154 14
12 11
13
16
15 Mirg
Figure 1 Several microRNA genes map within an imprinted locus on mouse distal 12 chromosome/human 14q32. The position of several imprinted genes is indicated by squares (Meg3 and protein-coding genes), vertical bars (small nucleolar RNA genes) or triangles (miRNA genes). Experimentally detected1 or predicted repeated gene copies of miRNAs (with hairpin structure ∼70 nt long; data not shown) are indicated by filled (red) or open triangles, respectively. Maternally expressed (red) and paternally expressed (blue) genes indicate the imprinting status. Several methylated regions are represented by circles (filled, hypermethylated; open, hypomethylated). In Mirg, spliced exons detected by genomic/cDNA comparison are represented by red boxes with dotted lines denoting splicing events. The maternally expressed antisense transcript to Rtl1 was initially suggested in sheep14 (called anti-PEG11). The schematic is not drawn to scale.
1LBME,
Centre National de la Recherche Scientifique (UMR 5099), IFR 109, Université P. Sabatier, 118 route de Narbonne, 31062 Toulouse Cedex, France. of Anatomy, University of Cambridge, Downing Street, Cambridge CB2 3DY, UK. 3Universität des Saarlandes, FR Genetik, Postfach 151150, D-66041 Saarbrücken, Germany. Correspondence should be addressed to J.C. (
[email protected]). 2Department
NATURE GENETICS VOLUME 34 | NUMBER 3 | JULY 2003
261
B R I E F C O M M U N I C AT I O N S
a
b
~1,228 ~6,000
~21–24
04 mir-127
anti-Rtl1 RNA 02
© 2003 Nature Publishing Group http://www.nature.com/naturegenetics
03
1,916 nt
Rtl1
mir-136
01
500
647 nt
Mirg Actb
P
S
S
K
Q
M
E
Rtl1
UCCAUCAUCAAAACAAAUGGAGU
3'
3'
AGGUAGUAGUUUUGUUUACCUCA
5'
miR-136
RT 1,665 nt
let-7
d 51
300
Gapd
miR-127 +4
c
2495
P
S
S
D
G
S D
Rtl1 AGCCAAGCUCAGACGGAUCCGA
Rtl1 RNA
3'
UCGGUUCGAGUCUGCCUAGGCU
3' 5'
miR-127 miR-136
mUPD12
P 1,916
647
RT 03 PCR
02 03 03 /02
02
1,665
RT
PCR
04 03 04/03
02 01 02/01
Figure 2 miR-127 and miR-136, two miRNAs expressed from the maternal chromosome, are antisense to Rtl1, a retrotransposon-like gene expressed only from the paternally inherited chromosome. (a) Reciprocal imprinting of Rtl1, mir-127 and mir-136. Total or poly(A)+ RNAs were isolated from normal embryos and embryos with maternal or paternal uniparental disomy for chromosome 12 (mUPD12 and pUPD12, respectively). Paternal allelic expression of Rtl1 imprinting is evident on northern blots of poly(A)+ RNAs using normal and UPD12 embryos (top). Maternal-specific expression of the miRNAs is indicated by northern blots for miR-127 (middle) or by primer extension for miR-136 (bottom). P, 32P-labeled primer. Probes for Gapd and let-7 were used as internal gel loading controls. Size is indicated in nucleotides. For unknown reasons, miR-136 could be detected only by primer extension. (b) Strand-specificity of maternally and paternally transcribed transcripts at Rtl1. Total RNA from mUPD12 or pUPD12 embryos was subjected to strand-specific RT–PCR with appropriate primers (depicted on top). The antisense RNA was detected only in mUPD12 RNA, whereas the sense RNA was detected only in pUPD12 RNA (bottom left). miR-127 and miR-136 are transcribed as a large precursor (bottom right). In each case, the primers used are indicated under the corresponding lanes (01–04; sequences available on request). RT, reverse transcriptase. (c) Mirg is maternally expressed in mouse embryos. Total RNA from normal, mUPD12 or pUPD12 embryos was reverse transcribed with primer Mirg8R. Subsequent PCR with primers Mirg7F and Mirg8R (primer sequences available on request) showed that Mirg was detected only in normal and mUPD12 RNA. Mirg7F and Mirg8R are located in the last two exons of the transcript (Fig. 1). As an additional control, RT–PCR (30 cycles) was done on the biallelically expressed gene Actb (encoding β-actin). RT, reverse transcriptase. (d) Schematic representation of miR-127 and miR-136 complementary sites in the Rtl1 open reading frame (nucleotide numbering begins at the start codon).
are not expressed (N.Y., S.-P.L. and A.C.F.-S., manuscript in preparation). Efforts are underway to decipher whether these miRNAs function in the regulation of Rtl1 gene expression during mouse development or in adults. The reciprocal imprinting of protein-coding genes and non-coding RNAs in imprinted clusters is a recurrent theme and suggests a regulatory role for non-coding RNAs in autosomal euchromatic epigenetic control8. Except for a direct role of the Air antisense RNA in the silencing of imprinted genes on mouse chromosome 17 (ref. 9), the involvement of RNA per se has not yet been shown. In contrast to plant miRNAs3,10, none of the metazoan miRNAs reported so far has perfect complementarity to any mRNA. Thus, the imprinted miR127 and miR-136 are the first miRNAs fully complementary to an endogenous cellular mRNA in animals, suggesting that the RNAi machinery could regulate this gene. RNA interference is considered a defense mechanism against mobile parasitic elements, including retrotransposons and retroviral-like sequences11; several imprinted genes share some features of these types of sequences12,13, notably the mouse gene Rtl1 identified as imprinted in this study. Thus, these findings identify a potential link between miRNAs, retrotransposon silencing and genomic imprinting. Animal care. All animal research was conducted in accordance with UK Government Home Office Licensing procedures. GenBank accession numbers. Mirg cDNA sequence, AJ517767; Rtl1, BK001261.
262
ACKNOWLEDGMENTS This work was supported by grants from the Toulouse Génopole, the Programme Interdisciplinaire du Centre National de la Recherche Scientifique Dynamique et Réactivité des Assemblages Biologiques and La Ligue contre le Cancer/Comité de Haute-Garonne (to J.C.) and by laboratory funds from the Centre National de la Recherche Scientifique and Université Paul-Sabatier, Toulouse (to J.-P.B.). Work in M.P.’s laboratory was supported by the Deutsche Forschungs Gemeinschaft. N.Y. is supported by a BBSRC studentship and S.-P. L. by a scholarship from the Taiwanese government. H.S. is supported by a PhD fellowship Allocation de Moniteur Normalien (Ministère de l’Education Nationale, de la Recherche et de la Technologie). COMPETING INTERESTS STATEMENT The authors declare that they have no competing financial interests. Received 11 March; accepted 18 April 2003 Published online 8 June 2003; doi:10.1038/ng1171 1. Lagos-Quintana, M. et al. Curr. Biol. 12, 735–739 (2002). 2. Cavaille, J., Seitz, H., Paulsen, M., Ferguson-Smith, A.C. & Bachellerie, J.P. Hum. Mol. Genet. 11, 1527–1538 (2002). 3. Llave, C., Xie, Z., Kasschau, K.D. & Carrington, J.C. Science 297, 2053–2056 (2002). 4. Hutvagner, G. & Zamore, P.D. Science 297, 2056–2060 (2002). 5. Reinhart, B.J. & Bartel, D.P. Science 297, 1831 (2002). 6. Volpe, T.A. et al. Science 297, 1833–1837 (2002). 7. Hall, I.M. et al. Science 297, 2232–2237 (2002). 8. Rougeulle, C. & Heard, E. Trends Genet. 18, 434–437 (2002). 9. Sleutels, F., Zwart, R. & Barlow, D.P. Nature 415, 810–813 (2002). 10. Rhoades, M.W. et al. Cell 110, 513–520 (2002). 11. Plasterk, R.H. Science 296, 1263–1265 (2002). 12. Neumann, B., Kubicka, P. & Barlow, D.P. Nat. Genet. 9, 12–13 (1995). 13. Yoder, J.A., Walsh, C.P. & Bestor, T.H. Trends Genet. 13, 335–340 (1997). 14. Charlier, C. et al. Genome Res. 11, 850–862 (2001).
VOLUME 34 | NUMBER 3 | JULY 2003 NATURE GENETICS
