| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Advances in Brief |
Center for Research on Occupational and Environmental Toxicology, Oregon Health Sciences University, Portland, Oregon 97201 [J. A. R., P. A. Y., M. S. T.]; Department of Pathology and Laboratory Medicine, University of Kentucky, Lexington, Kentucky 40536 [J. S.]; Department of Genetics, Rutgers, the State University of New Jersey, Piscataway, New Jersey 08854 [J. A. T.]; and Department of Cell Biology, Neurobiology, and Anatomy, University of Cincinnati, College of Medicine, Cincinnati, Ohio 45267-0521 [P. J. S.]
 |
ABSTRACT |
|---|
 Top ABSTRACT IntroductionMaterials and MethodsResultsDiscussionREFERENCES |
|---|
|
Introduction |
|---|
|
TopABSTRACTIntroductionMaterials and MethodsResultsDiscussionREFERENCES |
|---|
In previous work, we demonstrated methylation-associated silencing of mouse Aprt (5 , 6) , but did so in the context of a cell line hemizygous at Aprt. Therefore, all inactivation events were monoallelic. In the present study, we have isolated Aprt homozygous-deficient kidney cells directly from mice heterozygous for Aprt. These mice have a wild-type Aprt allele and a targeted allele bearing an insertion of the bacterial neo gene. We report that methylation-associated silencing of the wild-type Aprt allele was observed in some Aprt-deficient kidney clones. Interestingly, hypermethylation of the targeted allele and concomitant silencing of the inserted neo gene accompanied aberrant methylation of the wild-type allele. Because there was no selection for loss of neo expression from the targeted allele and because it did not occur in the absence of silencing of the wild-type allele, our data suggest a linkage between the two silencing events.
|
Materials and Methods |
|---|
|
TopABSTRACTIntroductionMaterials and MethodsResultsDiscussionREFERENCES |
|---|
4 weeks, DAP-resistant colonies were isolated and
expanded until permanent cell lines were established spontaneously. The
primary cells were grown in DMEM medium (Life Technologies, Inc.)
supplemented with 15% fetal bovine serum. The serum concentration was
lowered to 10% for the permanent cell lines. The mouse strains used
are described elsewhere (8)
.
Southern Blot Analysis.
Southern blot analysis for the Aprt gene upstream region has
been described previously (9)
.
Determination of Reversion Frequencies.
Aprt reversion frequencies were determined by plating the
deficient cell lines in medium supplemented with 50
µM azaserine (Sigma) and 60
µM adenine (Sigma; AzA medium). Because
Hprt expression interferes with this selection process (most
likely attributable to the presence of hypoxanthine in serum and/or as
a degradation product of adenine; data not shown),
Hprt-deficient subclones were selected with medium
supplemented with 10 µg thioguanine. Neo
reversion frequencies were determined by plating the cell lines in
medium supplemented with 500 µg/ml G418. The induction of revertants
with 3 µM 5-aza-dC (Sigma) was performed by
treating the cells for 48 h and then releasing them for 72
additional hours before plating in AzA or G418-supplemented media.
|
Results |
|---|
|
TopABSTRACTIntroductionMaterials and MethodsResultsDiscussionREFERENCES |
|---|
. The mutant frequencies for Aprt ranged from 2 to
20 x 10-5.
Sixteen DAP-resistant clones were expanded into permanent cell lines,
and all were found to lack APRT enzymatic activity (data not shown). A
preliminary molecular analysis demonstrated that seven cell lines
exhibited loss of heterozygosity for the wild-type allele, which is
most often attributable to mitotic recombinational events
(10)
, whereas the other nine retained this allele. A
sequence analysis for 7 of the presumptive point mutants (7-100, 8-38,
8-411, 8412, 10-26, 10-32, and 10-74) failed to identify bp
substitutions or any other type of intragenic alteration. This result
suggested epigenetic silencing of Aprt.
|
. The 2.5-
and 2.1-kb hybridization bands are observed normally at a ratio of 1:3.
This hybridization pattern indicates complete methylation of the
upstream H1 and H1a sites, partial methylation at a level of 25%
for the H2 site, and a complete lack of methylation of the H3 site
(Fig. 1)
. The H0 and H0a sites are also unmethylated in kidney DNA
(11)
. The 6-30 cell line, which exhibited loss of the
wild-type Aprt allele and is therefore homozygous for the
targeted allele, was found to have the normal methylation pattern (Fig. 2)
. This result demonstrates that the neo insertion did not
affect methylation of upstream H1a-H3 sites. Normal methylation
patterns were also observed for nonselected cell lines derived from
mice heterozygous and homozygous for Aprt (not shown). In
contrast, altered methylation patterns were observed for 8 of 9 nine
cell lines retaining the wild-type Aprt allele (Fig. 2
; data
not shown for 10-76) as evidenced by: (a) loss of the 0.4-
and 2.1-kb hybridization bands, (b) increased amounts of the
2.5-kb hybridization band relative to the 2.1-kb hybridization band,
and/or (c) higher molecular-sized hybridization bands. The
coincident loss of the 0.4- and 2.1-kb hybridization bands demonstrates
complete methylation of the H2 site. Appearance of hybridization bands
>2.5 kb is expected if methylation extends past the promoter region to
the H3 site. The methylation pattern for the 8-411 cell line appeared
normal.
|
62%. Such a result would appear like the hybridization pattern
observed for the 10-32 cell line, which reveals larger amounts of the
2.5-kb hybridization band relative to the 2.1-kb hybridization band.
Therefore, complete methylation at the H2 site demonstrates biallelic
methylation in the majority of the DAP-resistant cell lines retaining
the wild-type Aprt allele.
Spontaneous and Induced Reactivation of the Wild-type
Aprt Allele.
A hallmark of methylation-associated gene silencing in cultured cells
is its reversibility when the cells are exposed to the demethylating
agent 5-aza-dC (6)
. To determine if silencing of the
wild-type allele was reversible, five DAP-resistant cell lines (8-411,
8-412, 10-74, 10-76, and 6-30) were examined for spontaneous and
5-aza-dC-induced Aprt reactivation by measuring cloning
efficiencies in medium containing azaserine and adenine (AzA medium;
Table 1
). Growth in AzA medium requires functional Aprt expression.
Spontaneous reactivation of Aprt was observed in 3 of 5 cell
lines (8-412, 10-74, and 10-76) at frequencies ranging from 1.0 to
8.3 x 10-5. Treatment of these
cell lines with 5-aza-dC increased the reactivation frequencies from
40- to 250-fold consistent with elevated gene expression as a result of
a decrease in DNA methylation. Six of the 5-aza-dC-induced 8-412
revertants were examined by Southern blot analysis revealing lower
molecular weight hybridization bands indicative of decreased
methylation (Fig. 3)
. However, in these cases, demethylation did not appear to be
biallelic, (i.e., it did not restore the normal methylation
pattern), and novel-sized hybridization bands of 1.2- and 1.6-kb were
observed. Presumably, demethylation was detected for the reactivated
wild-type allele. Restoration of APRT enzymatic activity in revertants
selected in AzA medium was confirmed biochemically (data not shown).
|
|
. As expected, 5-aza-dC treatment had
no effect on the 6-30 cell line, which lacks a wild-type
Aprt allele. However, 5-aza-dC treatment induced revertants
for the 8-411 cell line, which exhibited a normal methylation pattern
at the time it immortalized spontaneously (Fig. 2)
. As described in
"Materials and Methods," the cells tested for reversion were
Hprt-deficient subclones of the original
Aprt-deficient primary clones. Thus, each subclone had
undergone a minimum of 30 additional population doublings before being
tested for 5-aza-dC-induced reactivation. Methylation of
Aprt was therefore evaluated in the 8-411 subclone that was
used for the reversion assay, and in this case, biallelic methylation
was observed (not shown). This result demonstrates that abnormal
methylation of both Aprt alleles in 8-411 occurred secondary
to the initial inactivation event. No change in methylation was
observed for the Hprt-deficient subclone of the 6-30 cell
line (not shown). Methylation for the Hprt-deficient
subclone of 8-412 was increased only slightly (Fig. 3)
when compared
with its parental cell line (Fig. 2)
.
Silencing of the neo Gene Insert-accompanied
Silencing of the Wild-type Aprt Allele.
The presence of a neo gene within the third exon of the
targeted allele (Fig. 1)
allowed us to determine whether methylation of
the upstream promoter region could influence expression of this insert.
Expression of the neo gene confers resistance to G418. The
original mass cultures and Hprt-deficient subclones (see
above) of four cell lines that exhibited epigenetic inactivation of
Aprt (8-411, 8-412, 10-74, and 10-76) were tested for G418
resistance, and all were found to be sensitive. In contrast, mass
cultures of five cells lines that did not contain a silenced wild-type
Aprt allele grew without apparent effect in the presence of
G418. One of these cell lines was 6-30 (Fig. 1)
, which had undergone a
loss of heterozygosity event as opposed to a silencing event. Two of
the cell lines were isolated without selection from mice that were
homozygous for the targeted allele, and the other two cell lines were
isolated without selection from mice that were heterozygous for the
targeted allele. Hprt-deficient subclones of the 6-30 cell
line and a heterozygous cell line termed KO6 (8)
were
tested and found to retain G418 resistance. In total, these results
demonstrate a correlation between epigenetic silencing of the wild-type
Aprt allele and silencing of the neo insert
within the targeted allele of the homologous chromosome.
To determine if the loss of neo expression was reversible,
and hence epigenetic, the four sensitive cell lines shown in Table 1
were tested for spontaneous and 5-aza-dC-induced neo
reactivation by selection in medium containing G418. The 6-30 cell line
was not used for these experiments because it was shown to be resistant
to G418 (see above). Spontaneous neo reactivation, as
demonstrated by the appearance of clones capable of growth in G418, was
observed for all four cell lines at frequencies ranging from
3.8 x 10-3
to 5.0 x 10-4. For three of four
cell lines (all except 8-411), 5-aza-dC treatment resulted in a
10-fold increase in the frequency of neo reactivation.
Expression of the bacterial neo gene inserted into targeted
allele (Fig. 1)
is under the control of a 278-bp promoter of the herpes
simplex TK gene. The last 130-bp of the promoter contain 15 CpG
dinucleotide pairs, suggesting that it could be sensitive to DNA
methylation, as has been shown for the mouse Aprt promoter
(5
, 6)
. Similar to the Aprt promoter, the TK
promoter contains Sp1 binding sites. To determine whether methylation
of the TK promoter correlated with expression, we performed a Southern
blot analysis for the G418-sensitive 8-412 cell line and for
spontaneous and 5-aza-dC- induced G418-resistant cells
(i.e., revertants) isolated from this cell line. For this
analysis, we took advantage of a HhaI site (Hh-1) located
immediately downstream of the second of two Sp1 binding sites within
the TK promoter, and we used an upstream XbaI site to
establish the 5' end of each hybridization fragment. A 205-bp fragment
beginning at the 5' end of the TK promoter and ending at the Hh-1 site
was used as the probe (Fig. 4A)
. As seen in Fig. 4B
, all HhaI sites
within the TK promoter and neo gene are methylated in the
8412 cell line, including the Hh-1 site located within the promoter, as
evidenced by the appearance of the 7.0-kb hybridization band.
Significant demethylation of the Hh-1 site, as evidenced by the
appearance of a 0.3-kb hybridization band, was observed in spontaneous
revertants and 5-aza-dC-induced revertants. Complete demethylation of
the Hh-1 site was observed in the induced revertants as compared with
partial demethylation in the spontaneous revertants. Both results most
likely represent a general decrease in methylation for the 15 CpG sites
that flank or are included within the two Sp1 binding sites in the TK
promoter, which contributes to reacquisition of neo
expression. Similarly, methylation status of the H3 (HpaII)
site located immediately downstream of the Aprt promoter
(Fig. 1)
has been shown to correlate with Aprt expression
(5
, 6)
. Complete methylation of the TK promoter and
neo gene was also observed in the DNA samples obtained from
the 10-74 and 10-76 cell lines and for the late passage 8-411 cell line
that was shown to have acquired biallelic methylation of the
Aprt promoter during continued passage in culture (see
previous section). Unfortunately, DNA from the 8-411 cell line at the
time it became immortalized and exhibited normal methylation at the
Aprt promoter region (Fig. 2)
is no longer available.
|
|
Discussion |
|---|
|
TopABSTRACTIntroductionMaterials and MethodsResultsDiscussionREFERENCES |
|---|
The experimental approach used in the present experiments was designed to capture in vivo events that result in loss of Aprt expression. Therefore, it is likely that the silencing events at Aprt observed in the kidney cells occurred in the animal. However, the timing of promoter region methylation is still an open question because a significant change in methylation was observed for one silenced cell line (8-411) during culture. Although high levels of DNA methylation have been noted in cultured cells (13) , it has not been demonstrated that actively transcribed promoters, as opposed to silenced promoters, can provide targets for de novo methylation events. Moreover, it is important to note that the promoter region methylation observed in this study is not simply an artifact of cell culture because it was not observed in the absence of silencing of the wild-type allele, nor was it observed in cells that contained only the targeted allele. Instead, if promoter region methylation is indeed a secondary event, it most likely targets an already silenced allele. Aberrant gene inactivation in vivo associated with hypermethylation has been demonstrated in malignant tissues (1) . DNA hypermethylation has also been detected in apparently normal tissues of the aging human colon (14) , suggesting that aberrant methylation is not restricted to malignant cell types.
A potential source of methylation-associated inactivation of Aprt is an upstream methylation center, which serves as a focus for methylation spreading and has been shown to induce inactivation when juxtaposed to Aprt promoters lacking a critical Sp1 binding site (15 , 16) . A model for methylation-associated gene inactivation in cancer attributable to spreading from heavily methylated repetitive DNA elements has been proposed (1 , 17) .
Methylation-associated silencing of the wild-type Aprt allele in the kidney cells was correlated with two additional and unexpected events. The first was hypermethylation of the targeted allele promoter region, indicative of biallelic methylation. As noted above, this result cannot be attributed solely to placement of the cells in culture, as it was not observed in the absence of epigenetic inactivation of the wild-type allele. It is also important to note that there was no selective pressure against the targeted allele because it does not produce a functional APRT protein. Instead, selection was directed solely against expression of the wild-type Aprt allele. Based on these observations, we suggest that methylation of the wild-type allele preceded methylation of the targeted allele. If so, the latter methylation event may be homology-dependent. Bestor and Tycko (18) have suggested a model for the transfer of methylation from one allele to its homologue via unresolved (i.e., nonrecombinant) Holliday junction intermediates. Homology-dependent inactivation in trans has been detected in plants, yeast, filamentous fungi, and Drosophila (19) . Alternatively, a cellular wide defect in the maintenance of appropriate methylation patterns that affected both Aprt promoters could have occurred in the selected kidney cells. Such a defect has been proposed as an explanation for methylation of multiple promoter regions in a given cancer cell (20) .
The second unexpected event that accompanied inactivation of the
wild-type allele was loss of G418 resistance. Such resistance requires
expression of a bacterial neo gene, in this case inserted
into the third exon of the Aprt targeted allele (Fig. 1)
.
Expression of this gene is driven by a herpes simplex TK promoter. The
simplest explanation for neo silencing is that it was linked
to methylation and/or inactivation of the upstream Aprt
promoter. As observed for the silenced wild-type Aprt
allele, restoration of neo expression was inducible by
treating the kidney cells with 5-Azc-dC. This implies that DNA
methylation also played a role in neo silencing. The
correlation between reacquisition of spontaneous or induced G418
resistance and demethylation of a HhaI site located in the
TK promoter confirms this expectation and further implicates
methylation of this promoter as playing an important role in
neo silencing. For the model of homology-dependent transfer
of methylation to be correct, methylation would have to spread from the
upstream Aprt promoter to TK promoter. Such spreading would
be required because no homologous region to TK/neo exists in
the wild-type Aprt allele.
In summary, we have reported biallelic methylation of mouse Aprt in normal kidney cells. These methylation events were associated with inactivation of Aprt and neo on the wild-type and targeted alleles, respectively. The absence of selection for neo inactivation suggests a potential role for allelic homology in this silencing event, although more global changes resulting in a trans effect are also possible. Further work will be necessary to elucidate the mechanisms involved in biallelic methylation and silencing, which are likely to have important implications for gene silencing in cancer.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
1 Supported by NIH Grants CA56383 and CA76528 (to
M. S. T.), PO 1 ES05652 (to P. J. S. and J. A. T.), and DK38185
(to J. A. T.). 
2 To whom requests for reprints should be
addressed, at CROET, L606, Oregon Health Sciences University, 3181 SW
Sam Jackson Park Road, Portland, OR 97201. Phone: (503) 494-2168; Fax:
(503) 494-6831; E-mail: turkerm{at}ohsu.edu
3 The abbreviations used are: DAP,
2,6-diaminopurine; Hprt, hypoxanthine phosphoribosyltransferase;
5-aza-dC, 5-aza-2'-deoxycytidine; TK, thymidine kinase; neo, neomycin;
AzA, azaserine and adenine medium.
Received 8/27/99. Accepted 5/ 9/00.
|
REFERENCES |
|---|
|
TopABSTRACTIntroductionMaterials and MethodsResultsDiscussionREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  D. Ronai, M. Berru, and M. J. Shulman The Epigenetic Stability of the Locus Control Region-Deficient IgH Locus in Mouse Hybridoma Cells Is a Clonally Varying, Heritable Feature Genetics, May 1, 2004; 167(1): 411 - 421. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. A. Yates, R. Burman, J. Simpson, O. N. Ponomoreva, M. J. Thayer, and M. S. Turker Silencing of Mouse Aprt Is a Gradual Process in Differentiated Cells Mol. Cell. Biol., July 1, 2003; 23(13): 4461 - 4470. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. S. Turker Autosomal mutation in somatic cells of the mouse Mutagenesis, January 1, 2003; 18(1): 1 - 6. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 O. N. Ponomareva, J. A. Rose, M. Lasarev, J. Rasey, and M. S. Turker Tissue-specific Deletion and Discontinuous Loss of Heterozygosity Are Signatures for the Mutagenic Effects of Ionizing Radiation in Solid Tissues Cancer Res., March 1, 2002; 62(5): 1518 - 1523. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. F Costello and C. Plass Methylation matters J. Med. Genet., May 1, 2001; 38(5): 285 - 303. [Abstract] [Full Text]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Cancer Research | Clinical Cancer Research |
| Cancer Epidemiology Biomarkers & Prevention | Molecular Cancer Therapeutics |
| Molecular Cancer Research | Cancer Prevention Research |
| Cancer Prevention Journals Portal | Cancer Reviews Online |
| Annual Meeting Education Book | Meeting Abstracts Online |
