HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

J. Biol. Chem., Vol. 275, Issue 48, 37915-37921, December 1, 2000

This Article Abstract Full Text (PDF) All Versions of this Article: 275/48/37915    most recent M004839200v1 Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Schumacher, A. Articles by Doerfler, W. Search for Related Content PubMed PubMed Citation Articles by Schumacher, A. Articles by Doerfler, W. Social Bookmarking                      What's this?

Epigenetic and Genotype-specific Effects on the Stability of de Novo Imposed Methylation Patterns in Transgenic Mice^*

Axel Schumacher, Paul A. Koetsier, Jennifer Hertz^§, and Walter Doerfler^¶

From the Institute of Genetics, University of Cologne, Weyertal 121, D-50931 Koeln, Germany

Received for publication, June 5, 2000, and in revised form, August 17, 2000

	ABSTRACT

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

The chloramphenicol acetyltransferase gene under the control of the late E2A promoter of adenovirus type 2 (Ad2) was introduced as transgene into the B6D2F1 mouse strain with mixed genetic background and became extensively de novo methylated. The methylation of this pAd2E2AL-CAT (7-1A) transgene was regulated in a strain-specific manner apparently depending on the site of integration. Transmission of the 7-1A transgene into an inbred DBA/2, 129/sv, or FVB/N genetic background led to a significant loss of methylation in the transgene, whereas C57BL/6, CB20, and Balb/c backgrounds favored the de novo methylation in very specific patterns. The newly established patterns of de novo methylation were transmitted to the offspring and remained stable for many generations, regardless of the heterozygosity of strain-specific DNA sequences present in these mouse strains. Segregation analyses showed a non-mendelian transmission of methylation phenotypes and suggested the involvement of dominant modifiers of methylation. The genotype-specific modifications of the transgene were followed for 11 backcross generations. These observations reflect an evolutionarily conserved mechanism directed against foreign, e.g. viral or bacterial, DNA at least in the chromosomal location of the 7-1A transgene. In seven additional mouse lines carrying the same transgene in different chromosomal locations, strain-specific alterations of methylation patterns were not observed.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Studies on the fate of foreign DNA introduced into an established mammalian genome are of considerable general interest. Upon integration, foreign (viral) DNA becomes frequently de novo methylated in very specific patterns (1-3) that are characterized by the gradual spreading of DNA methylation (4, 5). Following fertilization, similar epigenetic changes are known to affect a number of transgenes in mice as well as in transgenic fish and plants (6-12). The changes take place postzygotically in the early mouse embryo even before a stable somatic pattern is established (13). Already during early gastrulation, shortly after the genome-wide loss of methylation before the blastocyst stage, many DNA sequences become heavily remethylated (14-17).

In our laboratory, the adenovirus system has been used as a model to gain insight into consequences of foreign DNA integration. We have shown that the insertion of foreign DNA can lead to changes in cellular DNA methylation patterns even remote from the site of transgene integration (18, 19).

The regulation of de novo methylation of foreign DNA presumably involves a series of genes, sometimes referred to as modifier genes. Their putative function has frequently been ascribed to the demethylation of transgenes in DBA/2, FVB/N, and 129/sv backgrounds, whereas the Balb/c and C57BL/6 genotypes are often associated with extensive de novo methylation (20-25). Some transgenes show parent-of-origin effects that affect the genomic imprint at individual loci (for reviews see Refs. 26-28). The phenomenon of genomic imprinting at specific loci has first been demonstrated at randomly integrated transgene loci (8, 9, 29-32).

In this study, changes in the methylation of the transgene 7-1A (pAd2E2AL-CAT) have been investigated. This transgene has become methylated in a strain-specific manner in mice (12). We have studied the dynamics of 7-1A methylation for up to 12 generations in different mouse strains. We have also asked additional questions. Are the observed changes in transgene methylation restricted to specific 5'-CG-3' sites? How many factors influence the observed alterations in methylation? Is there a correlation between transgene methylation and the copy number of heterologous genome segments present?

	EXPERIMENTAL PROCEDURES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Mouse Strains-- The origins and relationships of the inbred mouse strains DBA/2, 129, FVB/N, C57BL/6, Balb/c, and CBA are reviewed in Beck et al. (33). A chart of inbred strain genealogies is available at the Mouse Genome Database.

Transgenic Mice-- The plasmid pAd2E2AL-CAT, which was used to generate the transgenic animals, was described previously (34). A complete map of the construct is shown in Fig. 2b. The male founder animal 7-1 was obtained after microinjection of unmethylated DNA. The 7-1 mouse line bearing two transgene arrays was subsequently separated into two substrains, 7-1A and 7-1B. Embryos used for microinjection resulted from matings between B6D2F₁ mice. 129/sv and CB20 mice were obtained from the mouse facility of the Institute of Genetics, University of Cologne, and all other strains (DBA/2, C57BL/6, Balb/c, FVB/N, and B6D2F₁) were received from Charles River, Sulzfeld, Germany.

DNA Extraction and Southern Blotting-- Animals were screened for the presence of the transgene by cutting tail tips of 3-4-week-old mice. Tail and organ tissues were frozen in liquid N₂, mortar-pulverized, and then incubated at 55 °C for a minimum of 12 h in lysis buffer (100 mM Tris-HCl, pH 8.5, 5 mM EDTA, 200 mM NaCl, 0.2% SDS, and 500 µg of proteinase K per ml). Proteinase K was then inactivated for 30 min at 65 °C. Subsequently, RNase A (20 µg/ml) was added, and incubation was continued for 1 h at 37 °C. The DNA was then purified by one phenol:chloroform:isoamyl alcohol (25:24:1) extraction followed by one phenol:chloroform (1:1) treatment. The DNA samples were ethanol-precipitated and resuspended in TE (10 mM Tris, pH 7.5, 1 mM EDTA, pH 8.0), and 20 µg of each sample was used to identify transgenic animals. The methylation levels were determined by cleaving the DNA with BamHI (10 units/µg DNA) for 5 h followed by either HpaII or MspI (10 units/µg DNA) incubation overnight. The fragments were separated by electrophoresis on 0.8% agarose gels in 1× TAE (40 mM Tris acetate, 1 mM EDTA, pH 8.0) for about 14 h at 30-40 V. Gels were stained with ethidium bromide (1 µg/ml) for 5-10 min, destained in water, and then prepared for downward Southern transfer (35) to Nylon plus membranes (Qiagen). The pAd2E2AL-CAT construct with flanking pBR322 vector sequences was used as ³²P-labeled hybridization probe in all hybridization experiments. The membranes were then exposed to x-ray films for 1-3 days. Methylation indices were determined by comparing the intensities of the HpaII (methylated) with the MspI (unmethylated) bands on the autoradiogram. The autoradiograms were scanned and analyzed by two different computer programs, Cybertech CS-1 Image Processing Software and TINA version 2.08. Some of the blots were also evaluated with the help of a PhosphorImager (Bio-Imaging Analyzer BAS 1000).

	RESULTS

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Establishing the 7-1A Mouse Line-- In earlier work in our laboratory, six mouse lines transgenic for the pAd2E2AL-CAT were established (34). The founder animal of line 7-1 originated from a microinjection experiment with an unmethylated transgene and had a mixed C57BL/6 × DBA/2 genetic background. Mouse line 7-1 was characterized by the stable integration of an estimated number of 10 transgene copies in a de novo methylated state. These animals had initially been mated to C57BL/6 animals for one generation. Out of this offspring, a transgenic male was selected to initiate systematic crossing experiments (Fig. 1, animal F₁, number 8).

View larger version (37K): [in this window] [in a new window]   Fig. 1.   Pedigree of 7-1 transgenic mice. All lines originated from a B6D2F1 hybrid genetic background. Transgenes were transmitted by heterozygous animals. After establishing a transgenic line, all further crosses were performed with heterozygous males. Non-transgenic animals were not included in this figure. BD indicated mating to non-transgenic B6D2F1 animals, B mating to C57BL/6, Ba mating to Balb/c, and D mating to DBA/2 animals. Open squares designate male and open circles designate female animals with 7-1A transgenes; asterisk indicates animals with demethylated transgenes (<30% methylated); filled squares indicate male, and filled circles indicate female animals with 7-1B transgenes; half-filled squares indicate male, and half-filled circles indicate female animals with 7-A/B transgenes. For comparisons of the methylation dynamics of the 7-1A transgenes, all mouse lines were derived from the same animal 16F6 with the exception of mouse line C57BL/6(b) that originated from animal 24F5.

The methylation patterns of all animals from the ensuing crosses were routinely screened by BamHI/HpaII restriction and subsequent Southern blot hybridization. Transgenic male animals were crossed further with non-transgenic hybrid strain B6D2F₁ females (C57BL/6 × DBA/2) for at least 4 generations. Although these transgenic animals were heterozygous for the pAd2E2AL-CAT transgene, they transmitted transgenes usually to about 70% of the offspring. Thus, the transgenes were located on separate chromosomes. As previously shown, the 7-1 mouse line carried two short arrays of four (A) and six (B) transgenes, respectively, in two different chromosomal locations (12). Starting from the F₄ cross, the separate transgene arrays were isolated on individual loci, resulting in the mouse lines 7-1A and 7-1B. Most of the six transgenes in the 7-1B animals were completely methylated at the 5'-CCGG-3' (HpaII) sites, whereas transmission of the 7-1A transgenes in a mixed B6D2F₁ background sometimes led to a significant decrease in transgene methylation. Males from both 7-1 mouse lines were subsequently mated to non-transgenic C57BL76, DBA/2, or Balb/c females. These crosses revealed that the methylation status of the 7-1A locus was influenced by the genetic background of the mouse strain. The DBA/2 background favored the loss of methylation from the 7-1A transgenes, whereas the C57BL/6 and Balb/c backgrounds not only maintained the methylation pattern but led to an increase in transgene methylation. Strain-specific effects were, however, not found for the 7-1B line. This finding indicated position effects on the methylation of the pAd2E2AL-CAT transgene. Starting with animals F₅ number 18 and F₅ number 24 (Fig. 1), consecutive matings with animals from different mouse strains (C57BL/6, Balb/c, DBA/2, CB20, FVB/N, or 129/sv) were performed to assess the effect of the inbred genetic background on the methylation patterns of the transgene in the 7-1A progeny.

DNA Methylation of the 7-1A Transgenes-- The breeding experiments generated four main 7-1A lines that were screened for over 11 generations as follows: two C57BL/6 lines with moderate to high de novo methylation of the transgene, very high methylation levels in the Balb/c line, and the decrease of methylation in the DBA/2 line. Typical transgene methylation profiles of the 7-1A transgene in animals from different strains are presented in Fig. 2a. Among different animals, there is some variability in the extent of transgene methylation. In the experiment shown here, densitometric evaluation of transgene methylation revealed 22% for the DBA/2 animals, 55% for C57BL/6, and 62% for Balb/c mice.

View larger version (66K): [in this window] [in a new window]   Fig. 2.   Methylation patterns of the 7-1A mouse lines. a, tail tip DNA (20 µg/lane) was cut with BamHI and the isoschizomeric endonuclease HpaII (H) or MspI (M) as indicated in the figure. Cleavage at the seven HpaII sites (see map in b) revealed a loss of methylation in the DBA/2 line, as evidenced by the almost complete absence of the 4.8- and 5.3-kbp off-size bands (arrows on the left) which represented the junctions linking the transgene to genomic DNA. The 3.3-kbp transgene repeat fragment in the C57BL/6 and Balb/c lines exhibited about double intensity when compared with DBA/2 DNA. This fragment was created by the cleavage at the unique BamHI restriction site. Densitometric evaluation (see "Experimental Procedures") of the cleavage patterns revealed that the transgene copies in DBA/2 mice were approximately 20% methylated, whereas the C57BL/6 lines (about 55% methylation) and Balb/c (about 62% methylation) favored de novo methylation of their transgenes. C, control animal DBA/2. b, the pAd2E2AL-CAT construct contained seven HpaII (H) restriction sites, one BamHI and a single HhaI site. The transgene consisted of two short pBR322 vector sequences at both ends, as well as a simian virus 40 (SV40) poly(A) sequence with the CAT gene as reporter gene, driven by the adenoviral E2A late (Ad2-E2AL) promoter. This construct was used as hybridization probe in all hybridization experiments.

The HpaII-MspI cleavage patterns were evaluated by densitometry. The ratio of HpaII bands representing methylated transgene segments, most of which were larger than 2.0 kbp,¹ to MspI fragments reflecting unmethylated transgene segments facilitated an assessment of the levels of 5'-CCGG-3' methylation in the transgenes of these mice. The cleavage patterns of the seven HpaII sites (Fig. 2b) revealed that the transgene in the DBA/2 line lost methylation, as evidenced by the almost complete absence of the 4.8- and 5.3-kbp off-size bands (Fig. 2a). These bands carried the junctions that linked the transgene to genomic DNA. Cleavage at the unique BamHI restriction site in the transgene produced a strong signal for the 3.3-kbp transgene repeat band, which exhibited about twice the intensity in the DNA from the highly methylating mouse strains C57BL/6 and Balb/c (Fig. 2a). The transgene copies with low levels of methylation were approximately 20% methylated, whereas the hypermethylated copies in the C57BL/6 strain reached levels of methylation between 40 and 60%. In the Balb/c genetic background, the methylation of the pAd2E2AL-CAT transgene was in excess of 80%. The high degree of 5'-CCGG-3' methylation correlated with the loss of CAT activity in the transgene. These methylation and expression patterns were identical in all organs, except when the transgenic construct had become demethylated and regained CAT activity (34). The increasing loss of methylation from the transgene in the germ line might signal an epigenetic reprogramming event at early stages of gametogenesis.

Maintenance and Variability of the Strain-specific Methylation in 7-1A Transgenes-- Obviously, in the chromosomal location of the 7-1A transgene, the genetic background of the animals played a major role in determining the maintenance or alterations of the methylation status in a foreign DNA transgene that had been introduced into the founder generation of hybrid C57BL/6-DBA/2 mice. The kind and number of genetic strain-specific factors affecting de novo methylation and its alterations in transgenes were unknown. Therefore, we performed a large number of crosses to transfer the transgene from the hybrid strain into a more homogeneous background and to keep it there for several generations. For example, starting with the F₇ animal 10 (Fig. 1) nine generations of backcrosses to DBA/2 mice resulted in a 99.8% DBA/2 genetic background. Since a DBA/2 background favored the loss of methylation of the 7-1A transgenes, matings were set up to investigate to what extent, after the prolonged crossing to DBA/2 animals, methylation in the transgene was decreased and whether the expression of the CAT gene could be reactivated. With successive backcrossings in a genetic DBA/2 background, methylation of the 7-1A transgene was reduced to and maintained at the 10-30% level (Fig. 3a). In these breedings, the transgene was always transmitted by the males to rule out a putative cytoplasmic reprogramming mechanism of the transgene during oogenesis. However, no significant variations in methylation were observed when transgenic 7-1A females were crossed with males of the same strain (data not shown). Regression analyses of methylation levels (Fig. 3a) demonstrated that in the DBA/2 animals transgene methylation remained stable at a level of 21%. Among siblings of the same litter, a 10% variation in transgene methylation was found. However, these differences were not transmitted to the offspring. Even in a pure genetic background of >99%, the transgene methylation remained mosaic and was associated with the silencing of the pAd2E2AL-CAT transgene.

View larger version (34K): [in this window] [in a new window]   Fig. 3.   Methylation status of foreign DNA in different inbred mouse strains. The pAd2E2AL-CAT transgene was crossed from a B6D2 hybrid strain into homogeneously inbred strains and was transmitted by heterozygous males. The resulting mouse lines were screened for up to 11 backcross generations. None of the analyzed mouse strains exhibited a correlation of transgene methylation and heterozygous DNA segments or strain-specific DNA sequences present (dash-double dot line). a, after crossing the 7-1A transgene into a DBA/2 background, transgene methylation remained at a low level. Although some variability in transgene methylation in different mice was found, the transgene methylation remained stable at a level of 21% (regression line, solid horizontal line) and was transmitted to the following generations. Open circles represent the methylation status of transgenic animals. Filled circle, founder animal. b, crossing the 7-1A transgene from a mixed genetic background ([C57BL/6 × DBA/2] × DBA/2)F1 into a Balb/c background led to the de novo methylation in the transgene. The methylation level then remained constant at around 61.5%. Each diamond represents the methylation status of one transgenic animal. Filled diamond, founder animal. c, each open circle represents a transgenic animal derived from founder animal 24F5 (mouse line C57BL/6-b, see Fig. 1), whereas a triangle represents an animal derived from founder animal 16F6 (mouse line C57BL/6-a). Filled triangle, founder animal. Both transgenic lines harbored hypermethylated 7-1A transgenes (fluctuating around 55%) after initial de novo methylation. The dashed line (dashes only) indicates the theoretical values for percentages of original DBA/2 (a), Balb/c (b), and C57LB/6 (c) genomes remaining after increasing numbers of crosses.

All transgenes in animals resulting from Balb/c × 7-1A-Balb/c or C57BL/6 × 7-1A-C57BL/6 crosses remained hypermethylated after the initial de novo methylation in the F₁ generation (Fig. 3, b and c). The transgenes in the C57BL/6 animals showed comparable methylation levels at around 55%; transgenic animals from backcrosses into Balb/c had a methylation level of 61.5% in the transgenes. In earlier work, we had shown in crosses with mixed genetic background BDF₁ × BDF₁ (Fig. 1) that in only 10% of the offspring transgene methylation was lost, and in the following generations the grandparental methylation patterns were usually reestablished (12). The ensemble of results (Fig. 3) suggests that both mouse strains, C57BL/6 and Balb/c, supply dominant factors favoring de novo methylation. The methylation dynamic of the 7-1A transgene resembles that of the TKZ751 transgene in mice (22, 23) in which the Balb/c genetic background led to stronger de novo methylation than the C57BL/6 background.

In addition to these B6 mouse lines, an additional C57BL/6-(b) line was established. In contrast to C57BL/6-(a), this line originated from animal number 24F₅ which was derived from animals showing no loss of methylation in their transgenes. After the first backcross, however, these animals exhibited a methylation pattern similar to that in the C57BL/6-(a) line. In both lines, the variation of transgene methylation was somewhat higher than in the other strains analyzed. For a more detailed analysis of these reproducibly observed strain-specific modifications, we backcrossed the 7-1A transgene into mouse strains FVB/N, 129/sv, or CB20 (data not shown). Starting from a DBA/2 animal with a hypomethylated transgene, its transfer into the FVB/N strain did not further affect the low level of transgene methylation. This observation paralleled the behavior of the RSVIgmyc transgenes in the same mouse strain (36). Similar results were obtained with 129/sv mice. Again like for other transgenes (22), the 129/sv background favored a moderate loss of methylation of the transgene. In testis, the transgene was fully demethylated, an effect typical for many strain-specifically methylated transgenes (8, 21, 34). In contrast to the FVB/N and 129/sv strains, the CB20 background led to de novo methylation. The CB20 mouse strain originated from Balb/c and C57BL/6 animals and might also harbor factors promoting methylation.

Stability of the 7-1A Methylation Patterns-- The observed strain-specific variations of transgene methylation are presumably subject to the control of modifying genes that are different in individual mouse strains. To test this hypothesis, a series of crosses was analyzed. At first, we investigated to what extent the highly methylated transgene phenotype in Balb/c mice could be reversed. Hence, we crossed the highly methylated transgene in the Balb/c animals into a DBA/2 background (Fig. 4). Conversely, the hypomethylated transgene from DBA/2 animals was backcrossed into a Balb/c background (Fig. 5a). In both experiments, the transgenes in the offspring showed the hypermethylated phenotype of the Balb/c mice. Therefore, the Balb/c genetic background seemed to play a dominant role in this process. At least after the first backcross, the methylation phenotypes were not immediately reversed.

View larger version (58K): [in this window] [in a new window]   Fig. 4.   Genetic dominance of the Balb/c background. Crosses of 7-1A Balb/c animals with non-transgenic DBA/2 mice always resulted in the maintenance of the hypermethylated state of the transgene (>50% methylated). Off- size band and 3.3-kbp transgene repeat bands were not cut. The same results were seen when transgenic DBA/2 mice were crossed with non-transgenic Balb/c animals (see Fig. 5a). The Balb/c genetic background seems to play a dominant role in the generation of the 7-1A methylation pattern. Tail DNA was cut with BamHI (lane 1) or with BamHI and HpaII (lanes 3 and 4). Double cleavage with BamHI and MspI (lane 2) served as reference for unmethylated fragments. Symbols are as follows: H, HpaII; M, MspI; B, BamHI. Open squares, male, non-transgenic animals; filled squares, male; filled circles, female mice carrying the 7-1A transgene.

View larger version (41K): [in this window] [in a new window]   Fig. 5.   Epigenetic stability of the hypomethylated transgene copies. a, methylation status of transgenic progeny from crosses with 7-1A DBA/2 males with hypomethylated transgene. The following crosses were analyzed: C57BL/6 × 7-1A-DBA/2 (black bars), Balb/c × 7-1A-DBA/2 (white bars), and DBA/2 × 7-1A-DBA/2 (dark gray bars). All male transgenic animals were randomly chosen for analyses and stemmed from the 9th DBA/2 backcross generation with an estimated 99.8% pure genetic background. All transgenes that were maintained in a DBA/2 background remained hypomethylated, whereas transmissions into Balb/c or C57BL/6 genetic backgrounds always led to significant de novo methylation of the transgenes. Therefore, both genetic backgrounds, Balb/c and C57/BL6, may contribute at least one dominant factor directing the de novo methylation of the 7-1A transgene. b, non-mendelian segregation of methylation phenotypes in progeny of crosses (C57BL/6 × DBA/2)F1 × 7-1A-DBA/2. One-third of the F1 animals exhibited the typical DBA/2 pattern (0-30% methylation), and two-thirds showed the C57BL/6 phenotype in their 7-1A transgenes (31-60% methylation). Methylation indices were determined as described under "Experimental Procedures."

We also tested whether the genetic factors would work in an independent fashion or, when inherited as heterozygous loci, by complementing each other. The hypermethylated transgene from Balb/c males was, therefore, crossed into a C57BL/6 genetic background. Intriguingly, all offspring showed the typical B6 methylation pattern of the 7-1A transgene (Fig. 6). These data indicated that the B6 allele was not only dominant over the DBA/2 allele (Fig. 5a) but also over the Balb/c allele. This result suggests the existence of a dominant factor affecting levels of DNA methylation in the C57BL/6 genetic background.

View larger version (70K): [in this window] [in a new window]   Fig. 6.   Factors affecting de novo methylation in C57BL/6 and Balb/c genetic backgrounds cannot complement each other. To test whether factors can work in an independent fashion or, when transmitted as heterozygous loci, by complementing each other, the hypermethylated 7-1A transgene was crossed from the Balb/c genetic background into the C57BL/6 background. Tail DNA (20 µg) of mice derived from the cross C57BL/6 × 7-1A-Balb/c was cut with BamHI and the methylation-sensitive endonuclease HpaII. The parent animal (p.Balb/c) exhibited a typical Balb/c methylation status of 73%. After the cross, all transgenes underwent a loss of methylation which led to the typical C57BL/6 methylation pattern (45-58% methylation). No increase in 5'-CCGG-3' methylation was found. Symbols are as follows: p, parental animal; H, HpaII; M, MspI; B, BamHI; open circles, female, non-transgenic animal; filled squares, male; filled circles, female mice carrying the 7-1A transgene. The arrows designate the positions of the 4.8- and 3.4 kbp fragments that decreased in intensities in the DNA from the offspring of this cross.

Segregation of the High and Low Methylation Phenotypes in the Progeny of 7-1A males with (B6 × D2)F₁ Females-- The initial strategy was to correlate the methylation in the offspring of crosses between DBA/2 males and (B6 × D2)F₁ females with previously characterized polymorphic loci. But instead of either the B6 or the D2 methylation phenotypes, additional methylation phenotypes were observed (data not shown). This result indicated that a multitude of factors could contribute to the establishment of distinct methylation phenotypes. Similarly complex patterns of methylation were encountered in segregation analyses with the transgenes E36 (25, 37, 38) and TKZ751 (22).

Further evidence for the complex role of genetic backgrounds in transgene methylation was adduced by crossing (C57BL/6 × DBA/2) F₁ females with 7-1A males of the DBA/2 background. Consistent with the notion of multiple methylation modifier loci, a non-mendelian segregation of methylation levels was apparent in the progeny (Fig. 5b). One-third of the offspring carried hypermethylated transgenes, whereas two-thirds of the progeny showed a de novo methylation that represented the B6 phenotype. These results indicate that the effect of several factors combine to modulate the transgene methylation in the 7-1A lines.

Interestingly, the observed de novo methylation might be directed to specific 5'-CG-3' dinucleotides (Fig. 7). HpaII and HhaI cleavage patterns of transgenic DNA from DBA/2, Balb/c, and C57BL/6 (not shown) were compared among animals of both sexes, which were randomly chosen from the sixth and seventh backcross generations. Although striking differences in HpaII-(CCGG) methylation were observed, the same animals showed no difference in their HhaI-(GCGC) cleavage patterns. This HhaI sequence is the only such site located within the pBR322 sequences of the construct. We conclude that the process of strain-specific transgene methylation seems to be a very specific modification that is dependent on more than one host-specific factor as well as on position effects.

View larger version (53K): [in this window] [in a new window]   Fig. 7.   The unique HhaI (GCGC) site escapes strain-specific loss of methylation. After HpaII treatment, the tail DNA was cleaved differentially, depending on the genetic background. The 3.3-kbp transgene repeat band was cut to almost 100% in transgenes in the DBA/2 (D) background (lane 3), whereas a large portion remained uncleaved in the Balb/c (Ba) genetic background (lane 4). When the same DNA was treated with HhaI, however, the cleavage patterns were not influenced by the genetic background (lanes 5 and 6). Symbols are as follows: H, HpaII; M, MspI; Hh, HhaI; D, DBA/2; C, control animal (DBA/2).

	DISCUSSION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Factors Affecting Transgene Methylation-- In recent years, a large number of transgenic mice have been generated to study various issues in genetics, like gene function or gene therapy. It is of considerable interest to understand the events accompanying or following the insertion of foreign DNA into established mammalian genomes (39). Not only the transgene affects the biology of the host, the transgene itself is often the target for several modifications. One of the major changes is the de novo methylation of the transgene upon insertion into the recipient genome. This heritable genetic modification modulates the overall genomic patterns of chromatin organization and gene expression without altering the nucleotide sequence. The extent of de novo methylation of the transgene depends on the mouse strain bearing the transgene. The 7-1A transgene used in this report consists of the CAT reporter gene driven by the adenovirus type E2A late promoter which is silenced after integration into the mouse genome. In early work from this laboratory, an inverse correlation between the methylation of the E2A late promoter and other adenovirus promoters and patterns of gene expression has been documented (2, 3, 34, 40, 41). The de novo methylation of transgenes is generally associated with altered chromatin organization that is inhibitory for transcription (6, 31). Nevertheless, the mechanisms behind this reprogramming remain an intriguing problem. Various factors are relevant in determining the extent and pattern of DNA methylation of the pAd2-E2AL-CAT transgene and of foreign DNA in general as follows: (a) position effects; (b) the mode of recombination; (c) specific bacterial or viral motifs in the foreign DNA; (d) the nucleotide sequence of the inserted construct; (e) the timing of insertion; and (f) the influence of other genes in the recipient cell or organism.

In this study, we have shown that the Ad2-E2AL-CAT transgene methylation is influenced in a strain-specific manner by a very coordinated regulation. In the mouse strains Balb/c, CB20, and C57BL/6, the 7-1A transgenes become highly de novo methylated, whereas a decrease in methylation is observed when the transgene is transmitted by DBA/2, 129/sv, or FVB/N animals. Obviously, the genetic background of the mouse strain can be a crucial factor in determining the expression levels of integrated foreign DNA. The expression of the transgene loci is inversely correlated with the extent of their methylation. Strain-specific effects are, of course, not restricted to our transgene model; other transgenes, regardless of their nature, can show similar characteristics when introduced into these mouse strains (Table I). The TKZ751 transgene (22) becomes repressed and methylated when TKZ751 males fertilize Balb/c eggs but not DBA/2 eggs. The brief transient exposure of C57BL/6 male pronuclei to the DBA/2 egg cytoplasm in the mouse compromises the ability of these nuclei in preimplantation development even when these nuclei are transplanted back into C57BL/6 recipient eggs (43). However, some transgenes are expressed in a strain-specific manner and are not de novo methylated upon passage into these backgrounds.

The Dynamics of 7-1A Methylation and the Role of Modifying Genes-- The dynamics of the de novo methylation of the 7-1A transgene might shed light on the amount and nature of strain-specific factors. Immediately after the first cross of the 7-1A transgene in the Balb/c or C57BL/6 backgrounds, de novo methylation was initiated. The newly established methylation pattern was subsequently transmitted to the offspring and remained stable in the following generations. This dominant methylation was a unique event with no further modifications, and indicated that both mouse strains supplied at least one, perhaps several, dominant allele that affected the stability of the pattern of methylation.

In 1991, Engler and co-workers (24) reported that a modifying locus regulated the methylation phenotype of a pHRD transgene that encoded a target sequence for the VDJ recombinase. Another interesting observation in the 7-1A lines was the decrease in methylation in the transgenes in the DBA/2, 129/sv, and FVB/N strains. Since these mice show no maintenance of the transgene methylation as compared with the methylation status in the founder animals, the loss of methylation might be an active process or the consequence of a failure in the maintenance mechanism. Thus, the methylation of the transgene could be considered the "default" state of the surrounding genomic sequences. Similar events were proposed for the RSVIgmyc transgene (36) which became hypomethylated and expressed in an inbred FVB/N genetic background but hypermethylated and silenced in inbred C57BL/6 mice. In contrast to the 7-1A construct, the RSVIgmyc transgene was also regulated by genomic imprinting, with the transgene being undermethylated when transmitted by the father and hypermethylated and silent when transmitted by the mother. Furthermore, the imprint was lost in the C57BL/6 genetic background but was restored by breeding the transgene into the FVB/N strain. Variation in the imprinting behavior in the C57BL/6 background suggested that elements in addition to the strain-specific factors had influenced de novo methylation during embryogenesis.

In mice carrying the 7-1A transgene, no significant variations in transgene methylation could be observed, when the transgene was transmitted by females instead of 7-1A males. Furthermore, the methylation pattern remained stable, when the transgene was transmitted as homozygous 7-1A locus (data not shown). These data suggest the following two alternative mechanisms affecting the methylation of the 7-1A transgene: an active decrease in methylation or methylation protection and a dominant mechanism supplied by the C57BL/6, Balb/c, and CB20 backgrounds resulting in the de novo methylation of the 7-1A transgene. The observation that the 7-1A transgene became dominantly de novo methylated in (B6 × D2)F₁ and (Balb/c × D2)F₁ hybrids was more compatible with the latter mechanism. However, the protection from de novo methylation would be incomplete and dosage-dependent. Successive backcrossings of the transgene into the DBA/2 background did not lead to the complete loss of methylation in the 7-1A transgene, although the decrease in transgene methylation could have been the result of changes in the proportion of cells carrying differentially methylated transgenes. In addition, the offspring of the highly inbred DBA/2 animals crossed with Balb/c animals always showed direct de novo methylation of the transgene.

The mechanism of de novo methylation might be related to an ancient defense system targeted against foreign DNA (1, 44-46). The 7-1A transgene consists of viral sequences (Ad2 and SV40) as well as bacterial DNA (pBR322; Fig. 2b). The importance of specific DNA sequences for the methylation status of transgenes has been previously discussed (36, 47). A short repetitive IgA sequence in the RSVIgmyc transgenes conveyed parental imprinting in a strain-specific manner. In another pHRD transgene, an E. coli sequence (gpr) might be recognized by the host and could therefore have served as a focus for the strain-specific modifier action with spreading of DNA methylation into the flanking sequences (47). It is unknown whether a specific sequence or repetitive structure within the 7-1A transgene can help trigger strain-specific methylation.

We have previously reported that the de novo methylation of the pAd2E2AL-CAT transgene is influenced by the position of integration (12). It is important to emphasize that only one of the six transgenic mouse lines then studied (34) showed strain-specific effects. These results are reminiscent of the phenomenon of position effect variegation in Drosophila where proximity to heterochromatic regions predisposes to variegation. Models have been proposed to explain the transgene inactivation by factors similar to the Drosophila Su(Var) genes that function in the epigenetic control of gene expression, presumably by modifying higher order chromatin structures (48). Recently, two vertebrate homologues, the human SUV39H1 and the mouse Suv39hl proteins, were isolated (49). The mammalian homologues contained highly conserved sequence motifs, the chromo and SET domains, a combination that was also preserved in the Schizosaccharomyces pombe silencing factor clr4 (50). Therefore, based on the notion of chromatin modeling and the fact that mammalian heterochromatin is often enriched in 5'-CG-3' methylation, the methylation status observed in the 7-1A transgene in different mouse strains might reflect secondary events following the formation of heterochromatin.

Our data also underscore the importance to choose the right mouse strain for gene targeting or embryogenic stem cell generation. The inbred genetic background may be decisive in determining the stability and continued expression of foreign DNA. Since the Balb/c and C57BL/6 backgrounds favor frequent inactivation of transgene constructs, mice from the DBA/2, FVB/N, or 129 strains would be preferable for targeting experiments.

Much further work will be necessary to understand the role of genetic background in the maintenance or alteration of the methylation status that affects the phenotypes associated with mutations in mice and in human disease. DNA methylation in mammalian genomes has frequently been interpreted as an epigenetic effect (51). Since patterns of DNA methylation in specific segments of a mammalian genome with a specific overall genetic background are stable and inherited, it might be more realistic to view these patterns as true genetic, rather than epigenetic, signals.

	FOOTNOTES

^* This work was supported in part by Deutsche Forschungsgemeinschaft Grant SFB274-A1, by the Center for Molecular Medicine Koeln, TV13, and EU Contract BMH4-CT96-0050.The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

Present address: Qiagen GmbH, Max-Volmer Strasse 4, D-40724 Hilden, Germany.

^§ Present address: 1135 Stanyan St., San Francisco, CA 94117. Supported by a fellowship of the Alexander von Humboldt Stiftung.

^¶ To whom correspondence should be addressed: Institute of Genetics, University of Cologne, Weyertal 121, D-50931 Koeln, Germany. Tel: 49-221-470-2386; Fax: 49-221-470-5163; E-mail: doerfler@scan. genetik.uni-koeln.de.

Published, JBC Papers in Press, August 22, 2000, DOI 10.1074/jbc.M004839200

	ABBREVIATIONS

The abbreviations used are: kbp, kilobase pairs; CAT, chloramphenicol acetyltransferase; Ad2, adenovirus type 2.

	REFERENCES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

CiteULike

Complore

Connotea

Del.icio.us

Digg

Reddit

Technorati

This article has been cited by other articles:

				C. Delaloy, J. Hadchouel, M. Imbert-Teboul, M. Clemessy, A.-M. Houot, and X. Jeunemaitre Cardiovascular Expression of the Mouse WNK1 Gene during Development and Adulthood Revealed by a BAC Reporter Assay Am. J. Pathol., July 1, 2006; 169(1): 105 - 118. [Abstract] [Full Text] [PDF]

				C. W. Park, B. T. Kren, D. A. Largaespada, and C. J. Steer DNA methylation of Sleeping Beauty with transposition into the mouse genome Genes Cells, August 1, 2005; 10(8): 763 - 776. [Abstract] [Full Text] [PDF]

				N. Suzuki, S. Imagawa, C. T. Noguchi, M. Yamamoto, and U. Klingmuller Do {beta}-globin, GATA-1,or EpoR regulatory domains specifically mark erythroid progenitors in transgenic reporter mice? Blood, November 1, 2004; 104(9): 2988 - 2989. [Full Text] [PDF]

				A. Schumacher and W. Doerfler Influence of in vitro manipulation on the stability of methylation patterns in the Snurf/Snrpn-imprinting region in mouse embryonic stem cells Nucleic Acids Res., March 5, 2004; 32(4): 1566 - 1576. [Abstract] [Full Text] [PDF]

				K. Inoue, N. Ogonuki, K. Mochida, Y. Yamamoto, K. Takano, T. Kohda, F. Ishino, and A. Ogura Effects of Donor Cell Type and Genotype on the Efficiency of Mouse Somatic Cell Cloning Biol Reprod, October 1, 2003; 69(4): 1394 - 1400. [Abstract] [Full Text] [PDF]

				W. DOERFLER, U. HOHLWEG, K. MULLER, R. REMUS, H. HELLER, and J. HERTZ Foreign DNA Integration--Perturbations of the Genome--Oncogenesis Ann. N.Y. Acad. Sci., September 1, 2001; 945(1): 276 - 288. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) All Versions of this Article: 275/48/37915    most recent M004839200v1 Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Schumacher, A. Articles by Doerfler, W. Search for Related Content PubMed PubMed Citation Articles by Schumacher, A. Articles by Doerfler, W. Social Bookmarking