- Email: [email protected]
The putative Drosophila methyltransferase gene dDnmt2 is contained in a transposon-like element and is expressed specifically in ovaries
Mechanisms of Development 95 (2000) 215±217
www.elsevier.com/locate/modo
Gene expression pattern
The putative Drosophila methyltransferase gene dDnmt2 is contained in a transposon-like element and is expressed speci®cally in ovaries Frank Lyko a, Allyson J. Whittaker a, b, Terry L. Orr-Weaver a, b, Rudolf Jaenisch a, b,* a
Whitehead Institute for Biomedical Research, 9 Cambridge Center, Cambridge, MA 02142, USA b Department of Biology, Massachusetts Institute of Technology, Cambridge, MA 02142, USA Received 17 February 2000; received in revised form 24 March 2000; accepted 27 March 2000
Abstract Several organisms, including Drosophila melanogaster, are apparently devoid of DNA methylation. This might re¯ect a highly restricted activity of DNA methyltransferases, a loss of methyltransferase activity during evolution or the dispensability of DNA methylation due to an ef®cient substitute mechanism. Vestiges of a Drosophila DNA methylation system have been identi®ed recently. We show here that the putative DNA methyltransferase gene, dDnmt2, is the component of a transposon-like element. This element also contains a second, novel open reading frame with homologies to a yeast protein involved in RNA processing. Both open reading frames are coordinately expressed and transcripts are present speci®cally in ovarian nurse cells as well as during early stages of embryonic development. q 2000 Elsevier Science Ireland Ltd. All rights reserved. Keywords: Dnmt2; Methyltransferase; Drosophila; dDnmt2; Imp4; Transposon; Ovaries; Nurse cells; Embryo
1. Results and discussion It has been shown recently, that the Drosophila genome contains a gene with signi®cant homologies to the putative mammalian DNA methyltransferase Dnmt2 (Hung et al., 1999; Tweedie et al., 1999). To characterize the genomic structure of the Drosophila dDnmt2 locus we made use of a fully assembled 30 kb genomic sequence that is available in the public databases (see Section 2 for accession number). This revealed that the dDnmt2 ORF is contained in a contiguous stretch of DNA that is not interrupted by introns. We also extended sequence analysis beyond dDnmt2 and identi®ed a second ORF that is very closely linked to dDnmt2 (Fig. 1A). Like dDnmt2, ORF2 is not interrupted by introns. ORF2 encodes a novel protein of 394 amino acids and matches to several Drosophila ESTs from an embryonic library (see materials and methods for accession numbers). The C-terminal region of ORF2 protein shows signi®cant homologies (50% amino acid similarity) to the yeast Imp4 protein (Lee and Baserga, 1999) (Fig. 1B). Imp4 is a component of ribonucleoprotein particles and has been shown to be necessary for rRNA processing (Lee and Baserga, 1999). The available information indicates that each of the two open reading frames is encoded by a single copy locus at * Corresponding author. Tel.: 11-617-258-5186; fax: 11-617-258-6505. E-mail address: [email protected] (R. Jaenisch).
33CD. Yet the compact genomic structure as well as the tight linkage of dDnmt2 and ORF2 were suggestive of a transposon-like element. Therefore we extended our sequence analysis further and investigated putative genes ¯anking dDnmt2 and ORF2. Alignment of genomic sequences with EST sequences revealed conventional exon/intron structures in the immediate vicinity of dDnmt2 and ORF2. We then looked for repetitive sequences ¯anking dDnmt2 and ORF2 and identi®ed a 12 bp inverted repeat close to the 3 0 ends of both open reading frames (Fig. 1A). This con®rmed the presence of another hallmark of transposon-like elements. We also analyzed the distribution of the dDnmt2 element among various Drosophila species by genomic Southern analysis under low stringency conditions. This revealed speci®c signals for dDnmt2 in D. melanogaster, D. teissieri and D. yakuba, but no signal for D. virilis and D. willistoni (Fig. 1C, upper panel). Hybridization of the same blot with a probe against the D. melanogaster rp49 locus (O'Connell and Rosbash, 1984) under high stringency conditions revealed distinct bands for all ®ve species tested. (Fig. 1C, lower panel). Taken together these results demonstrate a limited distribution of dDnmt2 among Drosophila species and thus reveal another similarity to mobile elements. To investigate the expression of dDnmt2, total RNA was isolated from various stages of D. melanogaster develop-
0925-4773/00/$ - see front matter q 2000 Elsevier Science Ireland Ltd. All rights reserved. PII: S 0925-477 3(00)00325-7
216
F. Lyko et al. / Mechanisms of Development 95 (2000) 215±217
Fig. 1. dDnmt2 is the component of a transposon-like element. (A) Structure of the dDnmt2 element. dDnmt2 is closely linked to a second open reading frame (ORF2). Both ORFs are separated by a 200 bp promoter region and are divergently transcribed from opposite strands. Arrows indicate directions of transcription. The element is ¯anked by a 12 bp inverted repeat (TTTTATTTTATC, black triangles). (B) Sequence alignment of ORF2 and S. cerevisiae Imp4. Identical amino acids are highlighted by black boxes, similar amino acids are highlighted by shaded boxes. (C) Limited distribution of dDnmt2 among Drosophila species. Genomic DNA was prepared from the Drosophila species indicated and analyzed by Southern blotting. Using low stringency hybridization conditions a signal for dDnmt2 could only be detected for D. melanogaster, D. teissieri, and D. yakuba (upper panel). A signal for the rp49 gene could be detected for all species tested, even under high stringency hybridization conditions (lower panel).
Fig. 2. Expression pattern of dDnmt2 and ORF2. (A) Developmental expression pattern. Total RNA was prepared from the developmental stages indicated and analyzed by Northern blotting. Speci®c transcripts were detected for dDnmt2 (about 1200 nucleotides length) and ORF2 (about 1400 nucleotides length). The roughly equal intensity of the rp49 signal indicates the loading of equal amounts of RNA. (B) Tissue-speci®c expression pattern in adult females. Total RNA was prepared from the tissues indicated and analyzed by Northern blotting. Speci®c transcripts were detected for dDnmt2 and ORF2. The roughly equal intensity of the rp49 signal indicates the loading of equal amounts of RNA. (C±E) Tissue-speci®c distribution of ORF2 transcript. ORF2 transcript was visualized by in situ hybridization with an ORF2 antisense riboprobe. (C) Stage 10 egg chamber with staining of the nurse cells (nc). (D) Stage 5 embryo with ubiquitous staining. (E) Stage 9 embryo with staining of neuroblasts (nb) and gut.
F. Lyko et al. / Mechanisms of Development 95 (2000) 215±217
ment and analyzed by Northern blotting. This showed a speci®c band of about 1200 nucleotides length in embryos and adult females (Fig. 2A). Our data are thus inconsistent with those of (Hung et al., 1999) who reported a larval-speci®c dDnmt2 RNA of 1650 nucleotides length. We cannot explain these discrepancies, but as our results were obtained with a cloned full-length dDnmt2 probe and were very well reproducible (see below), we are con®dent that the 1200 bp signal represents the genuine dDnmt2 RNA. We also analyzed expression of the second open reading frame by rehybridizing the same blot with a probe against ORF2. This revealed a speci®c band of about 1400 nucleotides length (Fig. 2A). The signal for ORF2 RNA was much stronger than that for dDnmt2 RNA, indicating a higher expression level. As for dDnmt2, signi®cant levels of ORF2 RNA were only detectable in embryos and in adult females (Fig. 2A) suggesting coordinate expression of dDnmt2 and ORF2. The female-speci®c expression pattern suggested that dDnmt2 and ORF2 RNA might be predominantly found in the ovaries of adult ¯ies. Northern analysis showed a signal of equal intensity for adult females and for ovaries, and no signal for females with ovaries removed (Fig. 2B). This demonstrated that dDnmt2 expression is restricted to ovaries and that the combined expression levels of all other adult female tissues are negligible. When the same blot was rehybridized with a probe against ORF2 we again observed strong signals that parallel the dDnmt2 expression pattern (Fig. 2B). This con®rms our previous conclusion of coordinate expression of dDnmt2 and ORF2. To determine the tissue-speci®c distribution of transcripts we performed RNA in situ hybridization. Levels of dDnmt2 RNA were too low to be detectable by this method (data not shown). Therefore we focused on the coordinately expressed ORF2 RNA. Ovaries were prepared and hybridized in situ to an ORF2 antisense riboprobe. This showed that the RNA is found predominantly in the nurse cells of stage 10 egg chambers (Fig. 2C). Nurse cells synthesize maternal RNAs that become stored in the developing oocyte. To analyze ORF2 RNA distribution during embryonic development we hybridized embryos of all developmental stages to the antisense riboprobe. This revealed the ubiquitous presence of ORF2 transcript in blastoderm stage embryos (Fig. 2D). During later stages, the distribution becomes more restricted and the RNA is found predominantly in neuroblasts and in the gut (Fig. 2E). Taken together, our results thus demonstrate that transcripts from the dDnmt2 element are particularly enriched during very early stages of embryonic development. It is presently unclear whether dDnmt2 encodes a functional methyltransferase because no activity could be detected in vitro (Tweedie et al., 1999) and no methylated DNA could be detected yet in the ¯y genome (Urieli-Shoval et al., 1982; Patel and Gopinathan, 1987). It is possible, however, that the Drosophila genome is methylated at a level below the current detection limit or that dDnmt2 has a different function which so far eluded characterization.
217
2. Materials and methods The complete dDnmt2 element and its neighboring genes are contained in the sequence AC019782. The accession numbers of ESTs matching to ORF2 are: AA202831, AA940987, AA942425, AI514749, AI515323, AA390378. The dDnmt2 ORF was ampli®ed from Oregon R genomic DNA by PCR using standard conditions. The cloned insert was veri®ed by sequencing and used as a probe in subsequent experiments. The ORF2 cDNA was obtained through the Berkeley Drosophila Genome Project (EST clone LD26635) and veri®ed by sequencing. Southern blots were hybridized overnight in 250 mM phosphate buffer and 7% SDS at 65 or 558C, as indicated. Blots were washed for 1 h in 40 mM phosphate buffer and 1% SDS at 65 or 558C, as indicated. For Northern blots, 6 mg of D. melanogaster (Oregon R) total RNA were separated on denaturing agarose gel and blotted using standard procedures (Sambrook et al., 1989). The size of RNAs was determined by comparison with an RNA size standard. RNA in situ hybridization on embryos and ovaries was performed as described previously (Royzman et al., 1999).
Acknowledgements We would like to thank Aki Hayashi-Hagihara for the rp49 probe and Anton Wutz for help with Northern blots. This work was supported by NIH grants GM39341 (to T.O.W.) and 5-R-35-CA44339 (to R.J.). A.W. was supported by an NIH predoctoral training grant. F.L. received a research fellowship from the Deutsche Forschungsgemeinschaft.
References Hung, M.S., Karthikeyan, N., Huang, B., Koo, H.C., Kiger, J., Shen, C.J., 1999. Drosophila proteins related to vertebrate DNA (5-cytosine) methyltransferases. Proc. Natl. Acad. Sci. USA 96, 11940±11945. Lee, S.J., Baserga, S.J., 1999. Imp3p and Imp4p, two speci®c components of the U3 small nucleolar ribonucleoprotein that are essential for pre18S rRNA processing. Mol. Cell. Biol. 19, 5441±5452. O'Connell, P.O., Rosbash, M., 1984. Sequence, structure, and codon preference of the Drosophila ribosomal protein 49 gene. Nucl. Acids Res. 12, 5495±5513. Patel, C.V., Gopinathan, K.P., 1987. Determination of trace amounts of 5methylcytosine in DNA by reverse-phase high-performance liquid chromatography. Anal. Biochem. 164, 164±169. Royzman, I., Austin, R.J., Bosco, G., Bell, S.P., Orr-Weaver, T.L., 1999. ORC localization in Drosophila follicle cells and the effects of mutations in dE2F and dDP. Genes Dev. 13, 827±840. Sambrook, J., Fritsch, E.F., Maniatis, T., 1989. Molecular Cloning: A Laboratory Manual, Cold Spring Harbor Laboratory Press, New York. Tweedie, S., Ng, H.H., Barlow, A.L., Turner, B.M., Hendrich, B., Bird, A., 1999. Vestiges of a DNA methylation system in Drosophila melanogaster? Nat. Genet. 23, 389±390. Urieli-Shoval, S., Gruenbaum, Y., Sedat, J., Razin, A., 1982. The absence of detectable methylated bases in Drosophila melanogaster DNA. FEBS Lett. 146, 148±152.
www.elsevier.com/locate/modo
Gene expression pattern
The putative Drosophila methyltransferase gene dDnmt2 is contained in a transposon-like element and is expressed speci®cally in ovaries Frank Lyko a, Allyson J. Whittaker a, b, Terry L. Orr-Weaver a, b, Rudolf Jaenisch a, b,* a
Whitehead Institute for Biomedical Research, 9 Cambridge Center, Cambridge, MA 02142, USA b Department of Biology, Massachusetts Institute of Technology, Cambridge, MA 02142, USA Received 17 February 2000; received in revised form 24 March 2000; accepted 27 March 2000
Abstract Several organisms, including Drosophila melanogaster, are apparently devoid of DNA methylation. This might re¯ect a highly restricted activity of DNA methyltransferases, a loss of methyltransferase activity during evolution or the dispensability of DNA methylation due to an ef®cient substitute mechanism. Vestiges of a Drosophila DNA methylation system have been identi®ed recently. We show here that the putative DNA methyltransferase gene, dDnmt2, is the component of a transposon-like element. This element also contains a second, novel open reading frame with homologies to a yeast protein involved in RNA processing. Both open reading frames are coordinately expressed and transcripts are present speci®cally in ovarian nurse cells as well as during early stages of embryonic development. q 2000 Elsevier Science Ireland Ltd. All rights reserved. Keywords: Dnmt2; Methyltransferase; Drosophila; dDnmt2; Imp4; Transposon; Ovaries; Nurse cells; Embryo
1. Results and discussion It has been shown recently, that the Drosophila genome contains a gene with signi®cant homologies to the putative mammalian DNA methyltransferase Dnmt2 (Hung et al., 1999; Tweedie et al., 1999). To characterize the genomic structure of the Drosophila dDnmt2 locus we made use of a fully assembled 30 kb genomic sequence that is available in the public databases (see Section 2 for accession number). This revealed that the dDnmt2 ORF is contained in a contiguous stretch of DNA that is not interrupted by introns. We also extended sequence analysis beyond dDnmt2 and identi®ed a second ORF that is very closely linked to dDnmt2 (Fig. 1A). Like dDnmt2, ORF2 is not interrupted by introns. ORF2 encodes a novel protein of 394 amino acids and matches to several Drosophila ESTs from an embryonic library (see materials and methods for accession numbers). The C-terminal region of ORF2 protein shows signi®cant homologies (50% amino acid similarity) to the yeast Imp4 protein (Lee and Baserga, 1999) (Fig. 1B). Imp4 is a component of ribonucleoprotein particles and has been shown to be necessary for rRNA processing (Lee and Baserga, 1999). The available information indicates that each of the two open reading frames is encoded by a single copy locus at * Corresponding author. Tel.: 11-617-258-5186; fax: 11-617-258-6505. E-mail address: [email protected] (R. Jaenisch).
33CD. Yet the compact genomic structure as well as the tight linkage of dDnmt2 and ORF2 were suggestive of a transposon-like element. Therefore we extended our sequence analysis further and investigated putative genes ¯anking dDnmt2 and ORF2. Alignment of genomic sequences with EST sequences revealed conventional exon/intron structures in the immediate vicinity of dDnmt2 and ORF2. We then looked for repetitive sequences ¯anking dDnmt2 and ORF2 and identi®ed a 12 bp inverted repeat close to the 3 0 ends of both open reading frames (Fig. 1A). This con®rmed the presence of another hallmark of transposon-like elements. We also analyzed the distribution of the dDnmt2 element among various Drosophila species by genomic Southern analysis under low stringency conditions. This revealed speci®c signals for dDnmt2 in D. melanogaster, D. teissieri and D. yakuba, but no signal for D. virilis and D. willistoni (Fig. 1C, upper panel). Hybridization of the same blot with a probe against the D. melanogaster rp49 locus (O'Connell and Rosbash, 1984) under high stringency conditions revealed distinct bands for all ®ve species tested. (Fig. 1C, lower panel). Taken together these results demonstrate a limited distribution of dDnmt2 among Drosophila species and thus reveal another similarity to mobile elements. To investigate the expression of dDnmt2, total RNA was isolated from various stages of D. melanogaster develop-
0925-4773/00/$ - see front matter q 2000 Elsevier Science Ireland Ltd. All rights reserved. PII: S 0925-477 3(00)00325-7
216
F. Lyko et al. / Mechanisms of Development 95 (2000) 215±217
Fig. 1. dDnmt2 is the component of a transposon-like element. (A) Structure of the dDnmt2 element. dDnmt2 is closely linked to a second open reading frame (ORF2). Both ORFs are separated by a 200 bp promoter region and are divergently transcribed from opposite strands. Arrows indicate directions of transcription. The element is ¯anked by a 12 bp inverted repeat (TTTTATTTTATC, black triangles). (B) Sequence alignment of ORF2 and S. cerevisiae Imp4. Identical amino acids are highlighted by black boxes, similar amino acids are highlighted by shaded boxes. (C) Limited distribution of dDnmt2 among Drosophila species. Genomic DNA was prepared from the Drosophila species indicated and analyzed by Southern blotting. Using low stringency hybridization conditions a signal for dDnmt2 could only be detected for D. melanogaster, D. teissieri, and D. yakuba (upper panel). A signal for the rp49 gene could be detected for all species tested, even under high stringency hybridization conditions (lower panel).
Fig. 2. Expression pattern of dDnmt2 and ORF2. (A) Developmental expression pattern. Total RNA was prepared from the developmental stages indicated and analyzed by Northern blotting. Speci®c transcripts were detected for dDnmt2 (about 1200 nucleotides length) and ORF2 (about 1400 nucleotides length). The roughly equal intensity of the rp49 signal indicates the loading of equal amounts of RNA. (B) Tissue-speci®c expression pattern in adult females. Total RNA was prepared from the tissues indicated and analyzed by Northern blotting. Speci®c transcripts were detected for dDnmt2 and ORF2. The roughly equal intensity of the rp49 signal indicates the loading of equal amounts of RNA. (C±E) Tissue-speci®c distribution of ORF2 transcript. ORF2 transcript was visualized by in situ hybridization with an ORF2 antisense riboprobe. (C) Stage 10 egg chamber with staining of the nurse cells (nc). (D) Stage 5 embryo with ubiquitous staining. (E) Stage 9 embryo with staining of neuroblasts (nb) and gut.
F. Lyko et al. / Mechanisms of Development 95 (2000) 215±217
ment and analyzed by Northern blotting. This showed a speci®c band of about 1200 nucleotides length in embryos and adult females (Fig. 2A). Our data are thus inconsistent with those of (Hung et al., 1999) who reported a larval-speci®c dDnmt2 RNA of 1650 nucleotides length. We cannot explain these discrepancies, but as our results were obtained with a cloned full-length dDnmt2 probe and were very well reproducible (see below), we are con®dent that the 1200 bp signal represents the genuine dDnmt2 RNA. We also analyzed expression of the second open reading frame by rehybridizing the same blot with a probe against ORF2. This revealed a speci®c band of about 1400 nucleotides length (Fig. 2A). The signal for ORF2 RNA was much stronger than that for dDnmt2 RNA, indicating a higher expression level. As for dDnmt2, signi®cant levels of ORF2 RNA were only detectable in embryos and in adult females (Fig. 2A) suggesting coordinate expression of dDnmt2 and ORF2. The female-speci®c expression pattern suggested that dDnmt2 and ORF2 RNA might be predominantly found in the ovaries of adult ¯ies. Northern analysis showed a signal of equal intensity for adult females and for ovaries, and no signal for females with ovaries removed (Fig. 2B). This demonstrated that dDnmt2 expression is restricted to ovaries and that the combined expression levels of all other adult female tissues are negligible. When the same blot was rehybridized with a probe against ORF2 we again observed strong signals that parallel the dDnmt2 expression pattern (Fig. 2B). This con®rms our previous conclusion of coordinate expression of dDnmt2 and ORF2. To determine the tissue-speci®c distribution of transcripts we performed RNA in situ hybridization. Levels of dDnmt2 RNA were too low to be detectable by this method (data not shown). Therefore we focused on the coordinately expressed ORF2 RNA. Ovaries were prepared and hybridized in situ to an ORF2 antisense riboprobe. This showed that the RNA is found predominantly in the nurse cells of stage 10 egg chambers (Fig. 2C). Nurse cells synthesize maternal RNAs that become stored in the developing oocyte. To analyze ORF2 RNA distribution during embryonic development we hybridized embryos of all developmental stages to the antisense riboprobe. This revealed the ubiquitous presence of ORF2 transcript in blastoderm stage embryos (Fig. 2D). During later stages, the distribution becomes more restricted and the RNA is found predominantly in neuroblasts and in the gut (Fig. 2E). Taken together, our results thus demonstrate that transcripts from the dDnmt2 element are particularly enriched during very early stages of embryonic development. It is presently unclear whether dDnmt2 encodes a functional methyltransferase because no activity could be detected in vitro (Tweedie et al., 1999) and no methylated DNA could be detected yet in the ¯y genome (Urieli-Shoval et al., 1982; Patel and Gopinathan, 1987). It is possible, however, that the Drosophila genome is methylated at a level below the current detection limit or that dDnmt2 has a different function which so far eluded characterization.
217
2. Materials and methods The complete dDnmt2 element and its neighboring genes are contained in the sequence AC019782. The accession numbers of ESTs matching to ORF2 are: AA202831, AA940987, AA942425, AI514749, AI515323, AA390378. The dDnmt2 ORF was ampli®ed from Oregon R genomic DNA by PCR using standard conditions. The cloned insert was veri®ed by sequencing and used as a probe in subsequent experiments. The ORF2 cDNA was obtained through the Berkeley Drosophila Genome Project (EST clone LD26635) and veri®ed by sequencing. Southern blots were hybridized overnight in 250 mM phosphate buffer and 7% SDS at 65 or 558C, as indicated. Blots were washed for 1 h in 40 mM phosphate buffer and 1% SDS at 65 or 558C, as indicated. For Northern blots, 6 mg of D. melanogaster (Oregon R) total RNA were separated on denaturing agarose gel and blotted using standard procedures (Sambrook et al., 1989). The size of RNAs was determined by comparison with an RNA size standard. RNA in situ hybridization on embryos and ovaries was performed as described previously (Royzman et al., 1999).
Acknowledgements We would like to thank Aki Hayashi-Hagihara for the rp49 probe and Anton Wutz for help with Northern blots. This work was supported by NIH grants GM39341 (to T.O.W.) and 5-R-35-CA44339 (to R.J.). A.W. was supported by an NIH predoctoral training grant. F.L. received a research fellowship from the Deutsche Forschungsgemeinschaft.
References Hung, M.S., Karthikeyan, N., Huang, B., Koo, H.C., Kiger, J., Shen, C.J., 1999. Drosophila proteins related to vertebrate DNA (5-cytosine) methyltransferases. Proc. Natl. Acad. Sci. USA 96, 11940±11945. Lee, S.J., Baserga, S.J., 1999. Imp3p and Imp4p, two speci®c components of the U3 small nucleolar ribonucleoprotein that are essential for pre18S rRNA processing. Mol. Cell. Biol. 19, 5441±5452. O'Connell, P.O., Rosbash, M., 1984. Sequence, structure, and codon preference of the Drosophila ribosomal protein 49 gene. Nucl. Acids Res. 12, 5495±5513. Patel, C.V., Gopinathan, K.P., 1987. Determination of trace amounts of 5methylcytosine in DNA by reverse-phase high-performance liquid chromatography. Anal. Biochem. 164, 164±169. Royzman, I., Austin, R.J., Bosco, G., Bell, S.P., Orr-Weaver, T.L., 1999. ORC localization in Drosophila follicle cells and the effects of mutations in dE2F and dDP. Genes Dev. 13, 827±840. Sambrook, J., Fritsch, E.F., Maniatis, T., 1989. Molecular Cloning: A Laboratory Manual, Cold Spring Harbor Laboratory Press, New York. Tweedie, S., Ng, H.H., Barlow, A.L., Turner, B.M., Hendrich, B., Bird, A., 1999. Vestiges of a DNA methylation system in Drosophila melanogaster? Nat. Genet. 23, 389±390. Urieli-Shoval, S., Gruenbaum, Y., Sedat, J., Razin, A., 1982. The absence of detectable methylated bases in Drosophila melanogaster DNA. FEBS Lett. 146, 148±152.