Biol. Chem. Hpppe-Seyler Vol. 372, pp. 1073-1079, December 1991
Distribution and DNA Methylation of a Repetitive Promoter Sequence Cloned from Mouse Embryonal Carcinoma Cells* Wolfgang A. SCHULZ, Maik S.W. OBENDORF and Thomas BEYER Institut für Physiologische Chemie I der Universität Düsseldorf
(Received 26 July/18 September 1991)
Summary: A 203 bp Hpall fragment (X9) cloned from F9 embryonal carcinoma cell DNA showed multiple bands on Southern hybridization towards mouse DNA digested with various restriction enzymes. No tandem repeats were evident. A 200 bp Mspl band corresponded to several copies of the cloned fragment and many elements appeared each associated with a 12-kb Pvull fragment. X9 is therefore a member of a moderately repetitive family probably dispersed in the mouse genome. X9 displayed polymorphisms between different Mus species, whereas multiple hybridizing bands were detected neither in DNA from rodent species outside the Mus
genus nor in DNA from man. Therefore, X9 provides a useful probe for studies of murine evolution. X9 contains a TATA box and a CCA AT box as well as several homologies to known transcription factor binding sites. When cloned in front of a reporter gene, it acted as a strong promoter in several different cell types. Most genomic sequences detected by X9 were highly methylated. A single copy appeared hypomethylated in adult mouse tissues, whereas several copies were hypomethylated in F9 and embryonic stem cells. It is speculated that X9 might represent the promoter of a murine gene family whose activity is regulated by DNAmethylation.
Verteilung und DNA-Methylierung einer aus embryonalen Mäuse-Karzinomzellen klonierten repetitiven Promotersequenz Zusammenfassung: Ein mit X9 bezeichnetes 203-bp///>flII-Fragment wurde aus F9 Embryokarzinomzellen kloniert und ergab bei der Southern-Hybridisierung gegen mit verschiedenen Restriktionsenzymen verdaute Maus-DNA jeweils mehrere Banden. Tandemwiederholungen wurden nicht gefunden. Eine 200-bp-Ms/?I-Bande entsprach offenbar mehreren Kopien des klonierten Fragments, und viele Elemente erschienen je mit einem 12-kb-PvwII-Fragment assoziiert. X9 gehört daher einer Familie mäßig repetitiver Elemente an, die im Mausgenom verteilt sind. X9 wies Polymorphismen zwischen verschiedenen Arten der Gattung Mus auf, wogegen weder in der DNA anderer Nagetiere noch des Menschen multiple Hybridisierungssignale gefunden wurden. X9
eignet sich daher als Probe zur Untersuchung der Mausevolution. X9 enthält ein TATA- und ein CCAAT-Element sowie weitere Homologien zu bekanntenTranskriptionsfaktorbindungsstellen. Vor ein Reportergen kloniert, zeigte es starke Promoteraktivität in verschiedenen Zelltypen. Der überwiegende Anteil der mit X9 hybridisierenden genomischen Sequenzen war hochmethyliert. In adulten Geweben war wahrscheinlich nur eine einzige Kopie untermethyliert, dagegen je mehrere Kopien in F9-Embryokarzinomzellen und in Embryonalen Stammzellen. X9 könnte die Promotersequenz einer Genfamilie im Mausgenom darstellen, deren Aktivität über DNAMethylierung kontrolliert wird.
Abbreviation: CAT, Chloramphenicol acetyltransferase (EC 3.2.1.28). * The sequence has since been submitted to the EMBL data bank and is accessible as X59818 (mouse, X9-DNA).
Brought to you by | Purdue University Libraries Copyright © by WalterAuthenticated de Gruyter & Co · Berlin · New York Download Date | 5/27/15 9:45 AM
1074
W. A. Schulz, M.S.W. Obendorf and T. Beyer
Vol. 372 (1991)
Key terms: DNA methylation, repetitive DNA elements, evolution, embryonal carcinoma cells, mouse.
A large part of the mammalian genome consists of multiple copies of specific sequences. Since the frequency and distribution of these repetitive elements differ among species, they can be used to study heredity and evolution111. In fact, there is evidence that repetitive elements contribute to evolutionary changes. Retrotransposons may move through the genome affecting the regulation and function of individual genes[2]. Rearrangements of centromeric tandem sequences may be important in the genetic separation of closely related species, e.g. within the Apodemus genus[3]. However, the function of most repetitive elements in somatic cells is unknown. One argument in favor of functional importance comes from the well-defined changes in DN A methylation that several major families of repetitive sequences undergo during murine development, e.g. IAP, LINE and minor satellite DNAs. In early embryonic cells, they are undermethylated, whereas they are strongly methylated in somatic tissue'4'6'. Repetitive sequences make up a significant fraction of the hypomethylated DNA in embryonic cells. Their hypomethylation may contribute to a distinct type of replication in early embryonic cells'7' and may allow transient expression of individual repetitive elements'2'6l Therefore, those sequences altering their methylation state during development are of particular interest. The present report describes the initial characterization of a new moderately repetitive sequence, X9, isolated from the hypomethylated fraction of embryonal carcinoma cell DNA. Materials and Methods High-molecular mass DNA from F9 mouse teratocarcinoma cells[sl was digested to completion with Hpall and separated on a 1.2% agarose gel. Fragments sized between 200 and 500 bp were ligated into the Clal site of pKS+ (Stratagene, Heidelberg), and transformed into a McrA, McrB positive XLl-blue strain of E. coli to select against internal methylation. Recombinant clones were identified by lac" selection and were chosen at random for further analysis. DNA was prepared from cell lines and tissues of laboratory mouse strains as described'9!. The particular DNA preparations from Apodemus species have been described131. DNAs from D3 129Sv mouse embryonic stem cells'10' were kindly provided by Dr. A. Gossler, Max-Delbrück-Laboratory, Köln, from Ellobius lutescens^ and hamsters by Dr. W. Vogel, University of Ulm, and from wild mice by Dr. H. Winking, Medical University of Lübeck. Restriction digestions, Southern blotting and DNA hybridization were performed as described'8·91. In-situ hybridization to metaphase chromosomes of mouse T-lymphoma cells was performed as described'121.
For the construction of X9-CATplasmids, the X9 insert was cloned into the Pstl and Sail sites of pGCATA and pGCATC'13' to yield pGCATX9A and pGCATX9C. Transient expression assays were performed by calcium phosphate mediated transfection as described'81 and CAT activity was determined by a standard method'141. All experiments were performed twice using duplicate plates for each plasmid. For pGCATX9Aand pGCATX9C, additional duplicate plates were included in two experiments with HepG2 cells which received O.lmM dibutyryl-cyclic AMP for the last 24 h before harvest.
Results and Discussion A series of clones was obtained from F9 mouse embryonal carcinoma cell DNA by random cloning of 200-500 bp Hpall fragments. Since Hpall sites are underrepresented in mammalian genomes and are only cut if the inner C of the recognition site is unmethylated, this fraction is enriched in hypomethylated DNA. About half the clones contained repetitive sequences as shown by hybridizing plasmid DNA to radiolabeled total mouse DNA or by direct Southern hybridization to mouse DNA. Southern hybridization established the cloned sequence X9 as moderately repetitive in the mouse genome. Multiple bands were seen with all restriction enzymes recognizing 4-, 5- and 6-bp sites without CpGs (Fig. 1). Considering the number of bands in the various digests and their relative intensities, the number of fragments per genome hybridizing to X9 is estimated as around 100. X9 was sequenced by the dideoxynucleotide method (Fig. 2) and found to be novel by comparison to the sequences stored in the EMBL data bank, Heidelberg (last: April 1991). X9 contains short homologies to several CpG-rich sequences, e.g. 69% within a45-bp stretch in the human spectrin mRNA. Sequences exhibiting such limited homologies are not expected to hybridize to X9 under stringent conditions and do therefore not contribute to the pattern seen on Southern blots. X9 itself is only moderately CpG-rich, with 6 CpGs in 203 bp and a OCPG of 0.3, whereas true CpG islands possess a 0.6[15]. Repetitive sequences dispersed in the genome, e.g. LINE elements, sometimes yield on Southern hybridization one distinctive prominent band with a par-
*The ratio observed/expected CpG is calculated as number of CpG x N(N = total number of nuÖCpG = number of C x number of G cleotides in the sequence being analysed).
Brought to you by | Purdue University Libraries Authenticated Download Date | 5/27/15 9:45 AM
1075
Mouse Repetitive Promoter Sequence
Vol. 372 (1991)
»_
X *-«
*Ξ ο ο ε Ξ ! +, ~ CL > α σι « Α - m
Φ +*
3
Ο
5 Τ»ο. α.> «ΛΟ ι
kb 2010642-
1.5-
o. kb H-
32 -
Fig. 1. Hybridization of X9 to mouse liver DNA digested with various restriction enzymes. 10 ^g of 129/Sv mouse liver DNA was digested with a 3-4 fold excess of the indicated restriction enzymes, separated on a 0.8% (A) or 1.5% (B) agarose gel, respectively, blotted and hybridized to X9 probe. The arrowhead in B denotes the position of a small Cfol fragment.
.5-
.1 Brought to you by | Purdue University Libraries Authenticated Download Date | 5/27/15 9:45 AM
1076
W. A. Schulz, M.S.W. Obendorf and T. Beyer
Vol. 372 (1991)
10 20 30 40 50 CCGGGTGCAC ACCTTAAATA CCCTACTGAC TCCTCCCAAT CACCCAGTCT GGCCCACGTG TGGAATTTAT GGGATGACTG AGGAGGGTTA GTGGGTCAGA
60 70 80 90 100 CATAACAGCA GTCCA2TGGT TCACAGCAAG TGATGACACC TCAGGCGAGC GTATTGTCGT CAGGTAACCA AGTGTCGTTC ACTACTGTGG AGTCCGCTCG
110 120 130 140 150 TGATGAGGCA TGGCGTCTGC GCAGTGCAGG CAGCTGGGAT GTGGCTAAGT ACTACTCCGT ACCGCAGACG CGTCACGTCC GTCGACCCTA CACCGATTCA
160 170 180 190 200 CTGAGGTAAT GAGGAAGTCA GGCGCAGGTC ATAAGACCTG GCAGCCATCC GACTCCATTk CTCCTTCAGT CCGCGTCCAG TATTCTGGAC CGTCGGTAGG
CGG GCC
Fig. 2. Sequence of X9. The sequence within the Clal cloning site as determined by sequencing both strands is shown. Restriction sites for Hpall strands and Pvull are indicated by double underlining, the putative TATA and CCAATbox and a possible cyclic AMP responsive element by single underlining.
B
Fig. 3. In-situ hybridization of X9 to mouse metaphase chromosomes. X9 insert (A) or plasmid vector (B) DNA was hybridized to metaphase spreads of TNYmouse T-lymphoma cells and exposed to photographic emulsion for 8 weeks.
Brought to you by | Purdue University Libraries Authenticated Download Date | 5/27/15 9:45 AM
Vol. 372 (1991)
Mouse Repetitive Promoter Sequence
ticular enzyme due to the conservation of restriction sites within unit repeats or within the flanking long terminal repeats of proviruses. Repetitive sequences arranged as tandem repeats, e.g. satellite DNA, usually show 'ladders' with one or several restriction enzymes that result from the occasional mutation of a restriction site present in the unit repeat. No clearcut tandem 'ladder' was generated by any restriction enzyme (Fig. 1). Digestion with Pvull, which cuts once within the cloned copy, resulted in most of the signal comprised in a 12-kb band (Fig. 1A), which would appear unusually large for a repeat resulting from two cuts within proviral long terminal repeats.The multiplicity of restriction fragments obtained overall (cf. Figs. 4 and 5) suggests that X9 represents a sequence located at several sites in the mouse genome. This assumption is supported by in-situ hybridization of X9 to mouse metaphase chromosomes, where specific signals were found distributed over several chromosomes (Fig. 3) Multiple bands were also found to hybridize following digestion with Mspl (which cuts CCGG, unless the outer C is methylated) and the patterns were nearly identical in DNA from embryonic cells and from adult tissue of 129/Sv mice (Fig. 4). Afundamentally different pattern was obtained with Hpall, the Mspl isoschizomer inhibited by CpG methylation. The majority of sequences hybridizing was of high molecular mass and was not resolved on standard agarose gels. In DNA from F9 cells and from the embryonic stem cell line D3, which is also derived from 129/ Sv mice^10', four additional bands reproducibly hybridized to X9 (Fig. 4). Of these, the bands at 250 bp, 2.1 and 2.5 kb, respectively, were not observed in adult tissues, whereas an approximately 200 bp Hpall fragment was reproducibly found in embryonic cells as well as in adult tissues, albeit weaker there (Fig. 4). It most probably corresponds to the cloned fragment. A 200 bp band was also detected in Mspl digests, but the signal was approximately 12 times stronger than the Hpall signal in F9 cells as quantified by densitometry (cf. Fig. 4).This suggests that the X9 fragment cloned initially is a representative of a distinct repetitive element. Almost all copies of this element appear to be methylated at their flanking Hpall sites in adult cells, while several ones are undermethylated in embryonic cells. The idea that most copies of X9 are highly methylated is corroborated by the results obtained with other restriction enzymes that are inhibited by cytosine methylation. A recognition site for Cfol (GCGC) is present in the cloned sequence (Fig. 2). As with Hpall, upon Cfol digestion essentially all sequences hybridizing to X9 were present in the high molecular
1077
mass fraction of adult tissue DNA (Fig. 1A). After very long exposures, a weak signal was detected at low size, possibly resulting from a single copy (Fig. IB). As is expected for DNA outside CpG islands, fragments resulting from digestions with other enzymes sensitive to CpG methylation (Clal, Haell, Neil, Notl, Sail, Smal, Sslll, Xhol) could not be resolved on standard agarose gels (not shown).
.5-
.1Fig. 4. Hybridization of X9 to tissue and embryonic cell DNA digested by Hpall or Mspl. 10 /x.g each of DNA from embryonal carcinoma (F9), embryonic stem cells (D3), or 129/Sv adult pancreas (129/Sv) was digested by either MspI (flanking lanes) or Hpall (inner lanes), separated on a 1.5% agarose gel and hybridized to X9 probe.
Brought to you by | Purdue University Libraries Authenticated Download Date | 5/27/15 9:45 AM
1078
W. A. Schulz, M.S.W. Obendorf and T. Beyer
ID U.
"ο χ
5-
Fig. 5. Species distribution of X9. 10 /-ig of tissue DNA from the indicated sources was digested by EcoRI, separated on an 0.8% agarose gel, blotted and hybridized to X9 probe. 129/Sv and AKR are laboratory strains, derived essentially from M. domesticus; M. domesticus and M. musculus 1,2 are individual wild mice from Northern Europe. Arrowheads denote polymorphisms between M. domesticus and M. musculus. Due to overexposure, the polymorphism at 6 kb denoted by the top arrowhead is not evident from the figure.
Brought to you by | Purdue University Libraries Authenticated Download Date | 5/27/15 9:45 AM
Vol. 372 (1991)
Mouse Repetitive Promoter Sequence
1079
of adjacent transcribed sequences. Cloning of adjacent sequences will also reveal whether all sequences hybridizing to X9 on Southern blots belong to the same family and how far their homology to the cloned copy extends. If X9 copies were indeed capable of functioning as strong promoters in the context of the mouse genome, the high methylation of most copies and the differential methylation of a few ones during development may indicate a role for DNA methylation in the control of X9 activity. We wish to thank several colleagues for providing DNA, Ms. /. Schaff ran for technical assistance, Dr. E.A. Komarova, Moscow, for help with some experiments, Dr. S. Adolph, Marburg and C. Klett, Ulm, for in-situ-hybridization data, Dr. M. Murphy and A. Klöne, Düsseldorf, for reading the manuscript, and Dr. H. Sies, Düsseldorf, for helpful suggestions and encouragement.This study was supported by the Deutsche Forschungsgemeinschaft (Schu 604/3-1).
X9A
X9C
SV2
Fig. 6. Transient transfection of X9-CATconstructs. Autoradiograph of the thin-layer plate from CAT assays using equal amounts of protein extract from HepG2 cells transfected in duplicate with 10 £ of plasmids pGCATA(A), pGCATX9A (X9A), pGCATX9C (X9C) or pSV2CAT(SV2), respectively.
References 1 2 3 4 5 6
cloned into chloramphenicol acetyltransferase expression vectors in two opposite orientations: pGCATX9A contains the sequences with the putative TATA box directed towards the CAT gene ('sense' orientation), and pGCATX9C away from it. In the first orientation, very high CAT activity was found after transient expression in HepG2 hepatoma cells, while the opposite orientation gave lower activity, albeit still significantly over background (Fig. 6). In neither orientation was the CAT activity affected by addition of O.lmM dibutyryl cyclic AMP (not shown). High promoter activity of the 'sense' orientation construct was also found in F9 embryonal carcinoma cells and in NIH3T3 fibroblasts. X9 is therefore capable of functioning as a promoter in many cell types upon transfection although proof for promoter function in situ will require the cloning
7 8 9 10 11 12 13 14 15 16 17
Hastie, N. (1989) in Genetic Variants and Strains of the Laboratory Mouse (Lyon, M.F. & Searle, A.G., eds.), 2nd edn., pp. 559-573, Fischer, Stuttgart. Di Nocera, P.P. & Sakaki, Y. (1990) Trends Genet. 6, 29-30. Hirning, U, Schulz, W. ., Just, W., Adolph, S. & Vogel, W. (1989) Chromosoma 98,450-455. Sanford, J.P., Chapman,V.M. & Rossant, J. (1985) Trends Genet, l, 89-93. Monk, M., Boubelik, M. & Lehnert, S. (1987) Development 99, 371 -382. Tolberg, M.E., Funderburk, S.J., Klisak, I. & Smith, S.S. (1987) J. Biol. Chem. 262,11167-11175. Selig, S., Ariel, M., Goiein, R., Marcus, M. & Cedar, H. (1988) EMBO J. 7, 419-426. Schulz, W.A. & Gais, G. (1989) Biochim. Biophys. Acta 1013,125-132. Schulz, W.A., Crawford, N. & Locker, J. (1988) Exp. Cell Res. 174, 433-447. Doetschmann,T.C., Eistetter, H., Katz, M., Schmidt, W. & Kemler, R. (1985)7. Embryol. Exp. Morphol. 87,27-45. Vogel, W., Steinbach, P., Djalali, M., Mehnert, K., Ali, S. & Epplen, J.T. (1988) Chromosoma 96,112-118. Adolph, S., Bartram, C.R. & Hameister, H. (1987) Cytogenet. Cell Genet. 44, 65-68. Frebourg,T. & Brison, O. (1988) Gene 65, 315-318. Gorman, C.M., Moffat, L.F. & Howard, B.H. (1982) Mol. Cell. Biol. 2,1044-1051. Gardiner-Garden, M. & Frommer, M. (1987) J. Mol. Biol 196,261-282. Pascale, E., Valle, E. & Furano, A.V. (1990) Proc. Natl. Acad. Sei. USA 87, 9481-9485. Roesler, W.J., Vandenbark, G.R. & Hanson, R.W. (1988) J. Biol. Chem. 263,9063-9066.
W.A. Schulz*, M.S.W. Obendorf and T. Beyer, Institut für Physiologische Chemie I, Heinrich-Heine-Universität, Moorenstr. 5,W-4000 Düsseldorf, Germany. * To whom correspondence should be addressed.
Brought to you by | Purdue University Libraries Authenticated Download Date | 5/27/15 9:45 AM
Brought to you by | Purdue University Libraries Authenticated Download Date | 5/27/15 9:45 AM
Distribution and DNA Methylation of a Repetitive Promoter Sequence Cloned from Mouse Embryonal Carcinoma Cells* Wolfgang A. SCHULZ, Maik S.W. OBENDORF and Thomas BEYER Institut für Physiologische Chemie I der Universität Düsseldorf
(Received 26 July/18 September 1991)
Summary: A 203 bp Hpall fragment (X9) cloned from F9 embryonal carcinoma cell DNA showed multiple bands on Southern hybridization towards mouse DNA digested with various restriction enzymes. No tandem repeats were evident. A 200 bp Mspl band corresponded to several copies of the cloned fragment and many elements appeared each associated with a 12-kb Pvull fragment. X9 is therefore a member of a moderately repetitive family probably dispersed in the mouse genome. X9 displayed polymorphisms between different Mus species, whereas multiple hybridizing bands were detected neither in DNA from rodent species outside the Mus
genus nor in DNA from man. Therefore, X9 provides a useful probe for studies of murine evolution. X9 contains a TATA box and a CCA AT box as well as several homologies to known transcription factor binding sites. When cloned in front of a reporter gene, it acted as a strong promoter in several different cell types. Most genomic sequences detected by X9 were highly methylated. A single copy appeared hypomethylated in adult mouse tissues, whereas several copies were hypomethylated in F9 and embryonic stem cells. It is speculated that X9 might represent the promoter of a murine gene family whose activity is regulated by DNAmethylation.
Verteilung und DNA-Methylierung einer aus embryonalen Mäuse-Karzinomzellen klonierten repetitiven Promotersequenz Zusammenfassung: Ein mit X9 bezeichnetes 203-bp///>flII-Fragment wurde aus F9 Embryokarzinomzellen kloniert und ergab bei der Southern-Hybridisierung gegen mit verschiedenen Restriktionsenzymen verdaute Maus-DNA jeweils mehrere Banden. Tandemwiederholungen wurden nicht gefunden. Eine 200-bp-Ms/?I-Bande entsprach offenbar mehreren Kopien des klonierten Fragments, und viele Elemente erschienen je mit einem 12-kb-PvwII-Fragment assoziiert. X9 gehört daher einer Familie mäßig repetitiver Elemente an, die im Mausgenom verteilt sind. X9 wies Polymorphismen zwischen verschiedenen Arten der Gattung Mus auf, wogegen weder in der DNA anderer Nagetiere noch des Menschen multiple Hybridisierungssignale gefunden wurden. X9
eignet sich daher als Probe zur Untersuchung der Mausevolution. X9 enthält ein TATA- und ein CCAAT-Element sowie weitere Homologien zu bekanntenTranskriptionsfaktorbindungsstellen. Vor ein Reportergen kloniert, zeigte es starke Promoteraktivität in verschiedenen Zelltypen. Der überwiegende Anteil der mit X9 hybridisierenden genomischen Sequenzen war hochmethyliert. In adulten Geweben war wahrscheinlich nur eine einzige Kopie untermethyliert, dagegen je mehrere Kopien in F9-Embryokarzinomzellen und in Embryonalen Stammzellen. X9 könnte die Promotersequenz einer Genfamilie im Mausgenom darstellen, deren Aktivität über DNAMethylierung kontrolliert wird.
Abbreviation: CAT, Chloramphenicol acetyltransferase (EC 3.2.1.28). * The sequence has since been submitted to the EMBL data bank and is accessible as X59818 (mouse, X9-DNA).
Brought to you by | Purdue University Libraries Copyright © by WalterAuthenticated de Gruyter & Co · Berlin · New York Download Date | 5/27/15 9:45 AM
1074
W. A. Schulz, M.S.W. Obendorf and T. Beyer
Vol. 372 (1991)
Key terms: DNA methylation, repetitive DNA elements, evolution, embryonal carcinoma cells, mouse.
A large part of the mammalian genome consists of multiple copies of specific sequences. Since the frequency and distribution of these repetitive elements differ among species, they can be used to study heredity and evolution111. In fact, there is evidence that repetitive elements contribute to evolutionary changes. Retrotransposons may move through the genome affecting the regulation and function of individual genes[2]. Rearrangements of centromeric tandem sequences may be important in the genetic separation of closely related species, e.g. within the Apodemus genus[3]. However, the function of most repetitive elements in somatic cells is unknown. One argument in favor of functional importance comes from the well-defined changes in DN A methylation that several major families of repetitive sequences undergo during murine development, e.g. IAP, LINE and minor satellite DNAs. In early embryonic cells, they are undermethylated, whereas they are strongly methylated in somatic tissue'4'6'. Repetitive sequences make up a significant fraction of the hypomethylated DNA in embryonic cells. Their hypomethylation may contribute to a distinct type of replication in early embryonic cells'7' and may allow transient expression of individual repetitive elements'2'6l Therefore, those sequences altering their methylation state during development are of particular interest. The present report describes the initial characterization of a new moderately repetitive sequence, X9, isolated from the hypomethylated fraction of embryonal carcinoma cell DNA. Materials and Methods High-molecular mass DNA from F9 mouse teratocarcinoma cells[sl was digested to completion with Hpall and separated on a 1.2% agarose gel. Fragments sized between 200 and 500 bp were ligated into the Clal site of pKS+ (Stratagene, Heidelberg), and transformed into a McrA, McrB positive XLl-blue strain of E. coli to select against internal methylation. Recombinant clones were identified by lac" selection and were chosen at random for further analysis. DNA was prepared from cell lines and tissues of laboratory mouse strains as described'9!. The particular DNA preparations from Apodemus species have been described131. DNAs from D3 129Sv mouse embryonic stem cells'10' were kindly provided by Dr. A. Gossler, Max-Delbrück-Laboratory, Köln, from Ellobius lutescens^ and hamsters by Dr. W. Vogel, University of Ulm, and from wild mice by Dr. H. Winking, Medical University of Lübeck. Restriction digestions, Southern blotting and DNA hybridization were performed as described'8·91. In-situ hybridization to metaphase chromosomes of mouse T-lymphoma cells was performed as described'121.
For the construction of X9-CATplasmids, the X9 insert was cloned into the Pstl and Sail sites of pGCATA and pGCATC'13' to yield pGCATX9A and pGCATX9C. Transient expression assays were performed by calcium phosphate mediated transfection as described'81 and CAT activity was determined by a standard method'141. All experiments were performed twice using duplicate plates for each plasmid. For pGCATX9Aand pGCATX9C, additional duplicate plates were included in two experiments with HepG2 cells which received O.lmM dibutyryl-cyclic AMP for the last 24 h before harvest.
Results and Discussion A series of clones was obtained from F9 mouse embryonal carcinoma cell DNA by random cloning of 200-500 bp Hpall fragments. Since Hpall sites are underrepresented in mammalian genomes and are only cut if the inner C of the recognition site is unmethylated, this fraction is enriched in hypomethylated DNA. About half the clones contained repetitive sequences as shown by hybridizing plasmid DNA to radiolabeled total mouse DNA or by direct Southern hybridization to mouse DNA. Southern hybridization established the cloned sequence X9 as moderately repetitive in the mouse genome. Multiple bands were seen with all restriction enzymes recognizing 4-, 5- and 6-bp sites without CpGs (Fig. 1). Considering the number of bands in the various digests and their relative intensities, the number of fragments per genome hybridizing to X9 is estimated as around 100. X9 was sequenced by the dideoxynucleotide method (Fig. 2) and found to be novel by comparison to the sequences stored in the EMBL data bank, Heidelberg (last: April 1991). X9 contains short homologies to several CpG-rich sequences, e.g. 69% within a45-bp stretch in the human spectrin mRNA. Sequences exhibiting such limited homologies are not expected to hybridize to X9 under stringent conditions and do therefore not contribute to the pattern seen on Southern blots. X9 itself is only moderately CpG-rich, with 6 CpGs in 203 bp and a OCPG of 0.3, whereas true CpG islands possess a 0.6[15]. Repetitive sequences dispersed in the genome, e.g. LINE elements, sometimes yield on Southern hybridization one distinctive prominent band with a par-
*The ratio observed/expected CpG is calculated as number of CpG x N(N = total number of nuÖCpG = number of C x number of G cleotides in the sequence being analysed).
Brought to you by | Purdue University Libraries Authenticated Download Date | 5/27/15 9:45 AM
1075
Mouse Repetitive Promoter Sequence
Vol. 372 (1991)
»_
X *-«
*Ξ ο ο ε Ξ ! +, ~ CL > α σι « Α - m
Φ +*
3
Ο
5 Τ»ο. α.> «ΛΟ ι
kb 2010642-
1.5-
o. kb H-
32 -
Fig. 1. Hybridization of X9 to mouse liver DNA digested with various restriction enzymes. 10 ^g of 129/Sv mouse liver DNA was digested with a 3-4 fold excess of the indicated restriction enzymes, separated on a 0.8% (A) or 1.5% (B) agarose gel, respectively, blotted and hybridized to X9 probe. The arrowhead in B denotes the position of a small Cfol fragment.
.5-
.1 Brought to you by | Purdue University Libraries Authenticated Download Date | 5/27/15 9:45 AM
1076
W. A. Schulz, M.S.W. Obendorf and T. Beyer
Vol. 372 (1991)
10 20 30 40 50 CCGGGTGCAC ACCTTAAATA CCCTACTGAC TCCTCCCAAT CACCCAGTCT GGCCCACGTG TGGAATTTAT GGGATGACTG AGGAGGGTTA GTGGGTCAGA
60 70 80 90 100 CATAACAGCA GTCCA2TGGT TCACAGCAAG TGATGACACC TCAGGCGAGC GTATTGTCGT CAGGTAACCA AGTGTCGTTC ACTACTGTGG AGTCCGCTCG
110 120 130 140 150 TGATGAGGCA TGGCGTCTGC GCAGTGCAGG CAGCTGGGAT GTGGCTAAGT ACTACTCCGT ACCGCAGACG CGTCACGTCC GTCGACCCTA CACCGATTCA
160 170 180 190 200 CTGAGGTAAT GAGGAAGTCA GGCGCAGGTC ATAAGACCTG GCAGCCATCC GACTCCATTk CTCCTTCAGT CCGCGTCCAG TATTCTGGAC CGTCGGTAGG
CGG GCC
Fig. 2. Sequence of X9. The sequence within the Clal cloning site as determined by sequencing both strands is shown. Restriction sites for Hpall strands and Pvull are indicated by double underlining, the putative TATA and CCAATbox and a possible cyclic AMP responsive element by single underlining.
B
Fig. 3. In-situ hybridization of X9 to mouse metaphase chromosomes. X9 insert (A) or plasmid vector (B) DNA was hybridized to metaphase spreads of TNYmouse T-lymphoma cells and exposed to photographic emulsion for 8 weeks.
Brought to you by | Purdue University Libraries Authenticated Download Date | 5/27/15 9:45 AM
Vol. 372 (1991)
Mouse Repetitive Promoter Sequence
ticular enzyme due to the conservation of restriction sites within unit repeats or within the flanking long terminal repeats of proviruses. Repetitive sequences arranged as tandem repeats, e.g. satellite DNA, usually show 'ladders' with one or several restriction enzymes that result from the occasional mutation of a restriction site present in the unit repeat. No clearcut tandem 'ladder' was generated by any restriction enzyme (Fig. 1). Digestion with Pvull, which cuts once within the cloned copy, resulted in most of the signal comprised in a 12-kb band (Fig. 1A), which would appear unusually large for a repeat resulting from two cuts within proviral long terminal repeats.The multiplicity of restriction fragments obtained overall (cf. Figs. 4 and 5) suggests that X9 represents a sequence located at several sites in the mouse genome. This assumption is supported by in-situ hybridization of X9 to mouse metaphase chromosomes, where specific signals were found distributed over several chromosomes (Fig. 3) Multiple bands were also found to hybridize following digestion with Mspl (which cuts CCGG, unless the outer C is methylated) and the patterns were nearly identical in DNA from embryonic cells and from adult tissue of 129/Sv mice (Fig. 4). Afundamentally different pattern was obtained with Hpall, the Mspl isoschizomer inhibited by CpG methylation. The majority of sequences hybridizing was of high molecular mass and was not resolved on standard agarose gels. In DNA from F9 cells and from the embryonic stem cell line D3, which is also derived from 129/ Sv mice^10', four additional bands reproducibly hybridized to X9 (Fig. 4). Of these, the bands at 250 bp, 2.1 and 2.5 kb, respectively, were not observed in adult tissues, whereas an approximately 200 bp Hpall fragment was reproducibly found in embryonic cells as well as in adult tissues, albeit weaker there (Fig. 4). It most probably corresponds to the cloned fragment. A 200 bp band was also detected in Mspl digests, but the signal was approximately 12 times stronger than the Hpall signal in F9 cells as quantified by densitometry (cf. Fig. 4).This suggests that the X9 fragment cloned initially is a representative of a distinct repetitive element. Almost all copies of this element appear to be methylated at their flanking Hpall sites in adult cells, while several ones are undermethylated in embryonic cells. The idea that most copies of X9 are highly methylated is corroborated by the results obtained with other restriction enzymes that are inhibited by cytosine methylation. A recognition site for Cfol (GCGC) is present in the cloned sequence (Fig. 2). As with Hpall, upon Cfol digestion essentially all sequences hybridizing to X9 were present in the high molecular
1077
mass fraction of adult tissue DNA (Fig. 1A). After very long exposures, a weak signal was detected at low size, possibly resulting from a single copy (Fig. IB). As is expected for DNA outside CpG islands, fragments resulting from digestions with other enzymes sensitive to CpG methylation (Clal, Haell, Neil, Notl, Sail, Smal, Sslll, Xhol) could not be resolved on standard agarose gels (not shown).
.5-
.1Fig. 4. Hybridization of X9 to tissue and embryonic cell DNA digested by Hpall or Mspl. 10 /x.g each of DNA from embryonal carcinoma (F9), embryonic stem cells (D3), or 129/Sv adult pancreas (129/Sv) was digested by either MspI (flanking lanes) or Hpall (inner lanes), separated on a 1.5% agarose gel and hybridized to X9 probe.
Brought to you by | Purdue University Libraries Authenticated Download Date | 5/27/15 9:45 AM
1078
W. A. Schulz, M.S.W. Obendorf and T. Beyer
ID U.
"ο χ
5-
Fig. 5. Species distribution of X9. 10 /-ig of tissue DNA from the indicated sources was digested by EcoRI, separated on an 0.8% agarose gel, blotted and hybridized to X9 probe. 129/Sv and AKR are laboratory strains, derived essentially from M. domesticus; M. domesticus and M. musculus 1,2 are individual wild mice from Northern Europe. Arrowheads denote polymorphisms between M. domesticus and M. musculus. Due to overexposure, the polymorphism at 6 kb denoted by the top arrowhead is not evident from the figure.
Brought to you by | Purdue University Libraries Authenticated Download Date | 5/27/15 9:45 AM
Vol. 372 (1991)
Mouse Repetitive Promoter Sequence
1079
of adjacent transcribed sequences. Cloning of adjacent sequences will also reveal whether all sequences hybridizing to X9 on Southern blots belong to the same family and how far their homology to the cloned copy extends. If X9 copies were indeed capable of functioning as strong promoters in the context of the mouse genome, the high methylation of most copies and the differential methylation of a few ones during development may indicate a role for DNA methylation in the control of X9 activity. We wish to thank several colleagues for providing DNA, Ms. /. Schaff ran for technical assistance, Dr. E.A. Komarova, Moscow, for help with some experiments, Dr. S. Adolph, Marburg and C. Klett, Ulm, for in-situ-hybridization data, Dr. M. Murphy and A. Klöne, Düsseldorf, for reading the manuscript, and Dr. H. Sies, Düsseldorf, for helpful suggestions and encouragement.This study was supported by the Deutsche Forschungsgemeinschaft (Schu 604/3-1).
X9A
X9C
SV2
Fig. 6. Transient transfection of X9-CATconstructs. Autoradiograph of the thin-layer plate from CAT assays using equal amounts of protein extract from HepG2 cells transfected in duplicate with 10 £ of plasmids pGCATA(A), pGCATX9A (X9A), pGCATX9C (X9C) or pSV2CAT(SV2), respectively.
References 1 2 3 4 5 6
cloned into chloramphenicol acetyltransferase expression vectors in two opposite orientations: pGCATX9A contains the sequences with the putative TATA box directed towards the CAT gene ('sense' orientation), and pGCATX9C away from it. In the first orientation, very high CAT activity was found after transient expression in HepG2 hepatoma cells, while the opposite orientation gave lower activity, albeit still significantly over background (Fig. 6). In neither orientation was the CAT activity affected by addition of O.lmM dibutyryl cyclic AMP (not shown). High promoter activity of the 'sense' orientation construct was also found in F9 embryonal carcinoma cells and in NIH3T3 fibroblasts. X9 is therefore capable of functioning as a promoter in many cell types upon transfection although proof for promoter function in situ will require the cloning
7 8 9 10 11 12 13 14 15 16 17
Hastie, N. (1989) in Genetic Variants and Strains of the Laboratory Mouse (Lyon, M.F. & Searle, A.G., eds.), 2nd edn., pp. 559-573, Fischer, Stuttgart. Di Nocera, P.P. & Sakaki, Y. (1990) Trends Genet. 6, 29-30. Hirning, U, Schulz, W. ., Just, W., Adolph, S. & Vogel, W. (1989) Chromosoma 98,450-455. Sanford, J.P., Chapman,V.M. & Rossant, J. (1985) Trends Genet, l, 89-93. Monk, M., Boubelik, M. & Lehnert, S. (1987) Development 99, 371 -382. Tolberg, M.E., Funderburk, S.J., Klisak, I. & Smith, S.S. (1987) J. Biol. Chem. 262,11167-11175. Selig, S., Ariel, M., Goiein, R., Marcus, M. & Cedar, H. (1988) EMBO J. 7, 419-426. Schulz, W.A. & Gais, G. (1989) Biochim. Biophys. Acta 1013,125-132. Schulz, W.A., Crawford, N. & Locker, J. (1988) Exp. Cell Res. 174, 433-447. Doetschmann,T.C., Eistetter, H., Katz, M., Schmidt, W. & Kemler, R. (1985)7. Embryol. Exp. Morphol. 87,27-45. Vogel, W., Steinbach, P., Djalali, M., Mehnert, K., Ali, S. & Epplen, J.T. (1988) Chromosoma 96,112-118. Adolph, S., Bartram, C.R. & Hameister, H. (1987) Cytogenet. Cell Genet. 44, 65-68. Frebourg,T. & Brison, O. (1988) Gene 65, 315-318. Gorman, C.M., Moffat, L.F. & Howard, B.H. (1982) Mol. Cell. Biol. 2,1044-1051. Gardiner-Garden, M. & Frommer, M. (1987) J. Mol. Biol 196,261-282. Pascale, E., Valle, E. & Furano, A.V. (1990) Proc. Natl. Acad. Sei. USA 87, 9481-9485. Roesler, W.J., Vandenbark, G.R. & Hanson, R.W. (1988) J. Biol. Chem. 263,9063-9066.
W.A. Schulz*, M.S.W. Obendorf and T. Beyer, Institut für Physiologische Chemie I, Heinrich-Heine-Universität, Moorenstr. 5,W-4000 Düsseldorf, Germany. * To whom correspondence should be addressed.
Brought to you by | Purdue University Libraries Authenticated Download Date | 5/27/15 9:45 AM
Brought to you by | Purdue University Libraries Authenticated Download Date | 5/27/15 9:45 AM
