In medium containing BrdU, BAR1 expression was further enhanced, associ-ated with more marked disruption of nucleosome positioning on the pro-moter region.. In this study, we examined wh
Trang 1genes with disruption of nucleosome positioning
Kensuke Miki1, Mitsuhiro Shimizu2, Michihiko Fujii1, Shinichi Takayama1,
Mohammad Nazir Hossain1and Dai Ayusawa1
1 Department of Genome System Science, Yokohama City University, Yokohama, Kanagawa, Japan
2 Department of Chemistry, Meisei University, Hino, Tokyo, Japan
Introduction
5-Bromodeoxyuridine (BrdU), which is frequently used
to measure DNA synthesis immunochemically in living
cells, is also well known to modulate various biological
functions when incorporated into DNA as
5-bromoura-cil instead of thymine Previously we have found that
BrdU very clearly induces a senescent-like phenomenon
in every type of mammalian cell and also in yeast cells
[1,2] Historically, BrdU has been used as a modulator
of cellular differentiation with cAMP and butyrate
[3,4] The latter two are found to target protein kinase
A and histone deacetylase, respectively, leading to func-tional understanding of cell signaling and gene expres-sion BrdU is thought to influence the expression of key genes involved in cellular differentiation However, the molecular mechanism underlying the actions of BrdU still remains a mystery in spite of many efforts [5]
We have extensively characterized genes up- or down-regulated by the addition of BrdU in various cell
Keywords
AT-tract; 5-bromodeoxyuridine; BAR1;
nucleosome positioning; transcriptional
derepression
Correspondence
D Ayusawa, Department of Genome
System Science, Yokohama City University,
Seto 22-2, Kanazawa-Ku, Yokohama,
Kanagawa 236-0027, Japan
Fax: +81 45 787 2193
Tel: +81 45 787 2193
E-mail: dayusawa@yokohama-cu.ac.jp
(Received 5 January 2010, revised 2 August
2010, accepted 2 September 2010)
doi:10.1111/j.1742-4658.2010.07868.x
5-Bromodeoxyuridine (BrdU) modulates the expression of particular genes associated with cellular differentiation and senescence when incorporated into DNA instead of thymidine (dThd) To date, a molecular mechanism for this phenomenon remains a mystery in spite of a large number of stud-ies Recently, we have demonstrated that BrdU disrupts nucleosome posi-tioning on model plasmids mediated by specific AT-tracts in yeast cells Here we constructed a cognate plasmid that can form an ordered array of nucleosomes determined by an a2 operator and contains the BAR1 gene as
an expression marker gene to examine BAR1 expression in dThd-auxotro-phic MATa cells under various conditions In medium containing dThd, BAR1 expression was completely repressed, associated with the formation
of the stable array of nucleosomes Insertion of AT-tracts into a site of the promoter region slightly increased BAR1 expression and slightly destabi-lized nucleosome positioning dependent on their sequence specificity In medium containing BrdU, BAR1 expression was further enhanced, associ-ated with more marked disruption of nucleosome positioning on the pro-moter region Disruption of nucleosome positioning seems to be sufficient for full expression of the marker gene if necessary transcription factors are supplied Incorporation of 5-bromouracil into the plasmid did not weaken the binding of the a2⁄ Mcm1 repressor complex to its legitimate binding site, as revealed by an in vivo UV photofootprinting assay These results suggest that BrdU increases transcription of repressed genes by disruption
of nucleosome positioning around their promoters
Abbreviations
BrdU, 5-bromodeoxyuridine; dThd, thymidine; MNase, micrococcal nuclease; TK, thymidine kinase.
Trang 2lines using PCR-based cDNA subtractive hybridization
and DNA microarray analysis [6,7] Such
BrdU-responsive genes behave similarly in normal human
fibroblasts undergoing replicative senescence BrdU
decondenses particular regions of chromosomes after
incorporation into DNA [8], suppresses position effect
variegation [9] and restores expression of silenced
genes [10] Consistent with this, BrdU-responsive genes
are located on particular regions of human
chromo-somes, forming clusters on or nearby Giemsa-dark
bands of human chromosomes [11,12] AT-tract minor
groove binders, such as distamycin A, netropsin,
Hoe-chst 33258 and the AT-hook protein HMG-I, have all
been shown to markedly potentiate the effects of BrdU
[11,13] On the basis of the above observations, we
suggest that BrdU targets certain types of AT-rich
sequence and alters the chromatin structure to induce
particular genes
In eukaryotes, DNA fibers exist as regularly arrayed
beads of nucleosomes Nucleosomes restrict the
acces-sibility of transcription factors to promoters and
regu-latory sequences of genes Thus, an alteration of
nucleosome positioning is an essential step for the
transition from a repressed state to an active state by
aid of chromatin remodeling complexes [14–16] In our
previous studies, we addressed the effect of BrdU on
nucleosome positioning in vivo using TALS plasmids
[17], which have been successfully utilized to study
nucleosome positioning in Saccharomyces cerevisiae
We have clearly shown that 5-bromouracil
incorpo-rated into the plasmids disrupts nucleosome
position-ing by inducposition-ing A-form-like DNA conformation in
yeast cells [18] In mammalian cells, histones were
shown to bind more tightly to 5-bromouracil-substi-tuted DNA in vitro than normal DNA containing thy-mine [19,20]
In this study, we examined whether BrdU induces the expression of genes associated with the destabiliza-tion of nucleosome posidestabiliza-tioning with model plasmids containing the BAR1 gene as a marker gene We showed that BrdU increases the transcription of repressed genes by disruption of nucleosome position-ing around their promoters These data will facilitate the understanding of the role of BrdU and AT-tracts
in the induction of particular genes
Results
Construction of model plasmids
In yeast MATa cells, the a2⁄ Mcm1 repressor binds to the a2 operator and acts to repress MATa cell-spe-cific genes such as BAR1, STE2 and STE6 [21–25] Previously, our high-resolution mapping of micrococ-cal nuclease (MNase) cleavage sites has indicated that nucleosomes were well positioned around the pro-moter region of the genomic BAR1 gene [26] and such nucleosome positioning was required for its full repression [27] Here, we constructed a plasmid pRS-BAR1 that contains the pRS-BAR1 gene as a model sys-tem (Fig 1) to easily examine the role of BrdU in gene expression and nucleosome positioning in vivo
To further examine the effects of AT-tracts on nucle-osome positioning, we inserted them into a KpnI site adjacent to the TATA box These plasmids were introduced into a thymidine (dThd)-auxotrophic
Inserts at the Kpn I site pRS001-BAR1 pRS801-BAR1
: T3CCT6CT5GCT5CT7 : A34
BAR1 -UTR -UTR pRS-BAR1
TATA
2 op
mRNA Kpn I
+1 130
– 219 – 66 +95 +255
I II III IV V
60 –500
Xho I
– 235 – –
+420 –158
EcoRV
+1667
BAR1
Fig 1 Schematic representation of pRS-BAR1 and its derivatives Plasmid pRS-BAR1 contains a genomic fragment spanning the BAR1 gene regulated by a2 ⁄ Mcm1 repressor The a2 operator (a2 op), the TATA box (TATA) and the BAR1 coding sequence are denoted by hatched, filled and dotted boxes, respectively The transcriptional initiation site is indicated by a bent arrow The site of KpnI, to which AT-tracts are inserted, is indicated by a vertical line The probes A and B are indicated by gray boxes Positions of stable nucleosomes are shown by shadowed ellipses with numbers.
Trang 3mutant of yeast cells to ensure quantitative
incorpora-tion of BrdU into DNA [2]
Nucleosome positioning on pRS-BAR1
We first examined whether pRS-BAR1 forms a stable
and precise array of nucleosomes in dThd medium
using the indirect end-labeling method DNA samples
were completely digested with XhoI after partially
digestion with MNase, and were subjected to Southern
blot analysis with probe A Five MNase cleavage sites
indicated by *a, *b, *c, *d and *e were observed in the
naked DNA sample, and these sites were protected in
a chromatin sample of pRS-BAR1 (Fig 2A, lanes 1
and 2) Also, the cleavage sites showed equal intervals
(140–150 bp) in the chromatin sample (Fig 2A, lane
1) These results showed that at least five stable
nucle-osomes were precisely and stably formed from the a2
operator to the coding region of the BAR1 gene on
pRS-BAR1 in vivo (Fig 1, bottom panel)
We then examined how BrdU affects nucleosome
positioning on the pRS-BAR1 plasmid When the
MATa cells were cultured in BrdU medium, the two
bands (*b and *c) that were protected in the chromatin
sample prepared in dThd medium were evident in the
chromatin sample prepared in BrdU medium (Fig 2A,
lanes 1 versus 3) Densitometrical measurement clearly
showed these differences (Fig 2B) Consistently, the
band corresponding to the linker region between
nucle-osomes II and III became broader in the same sample
(Fig 2A, lane 3) Because there was no significant
dif-ference in MNase cleavage patterns between naked
DNA samples prepared in dThd and BrdU medium,
BrdU did not affect the specificity and sensitivity to
MNase (Fig 2A, lanes 2 versus 4) These results
indi-cate that BrdU destabilized nucleosome positioning
We then examined the sequence in nucleosome II, and
found the following: the AT-content of the sequence
was 69%, whereas in other nucleosomes it was
approx-imately 60%, and An, Tn or (AT)n tracts (n‡ 6) were
found in six sites in nucleosome II, but not in
nucleo-some III Therefore, BrdU seemed to destabilize the
positioning of nucleosome II through AT-rich
sequences in nucleosome II, which then led to the
destabilization of nucleosome III The same results
were obtained when nucleosome positioning was
deter-mined from the EcoRV site with probe B (Fig 1, S1,
lanes 1–4)
Effects of AT-tracts on nucleosome positioning
We examined two derivatives of pRS-BAR1,
pRS001-BAR1 and pRS801-pRS001-BAR1, containing T3CCT6CT5
GCT5CT7 and A34, respectively, in a promoter region
of the BAR1 gene (Fig 1A) Insertion of these AT-tracts did not affect MNase cleavage patterns in naked DNA samples (Fig 2A, even-numbered lanes) The MNase cleavage pattern in a chromatin sample of pRS001-BAR1 was similar to that in the chromatin sample of pRS-BAR1 in dThd medium, although the bands *c and *d were more evident in the former sam-ple (Fig 2A, lanes 1 versus 5) This suggests that the positioning of nucleosomes I and II on pRS001-BAR1 was less stable than on pRS-BAR1 in dThd medium The bands *b, *c and *e in the chromatin sample of pRS001-BAR1 were more marked in BrdU medium than in dThd medium (Fig 2A, B, lanes 5 versus 7), although the MNase cleavage patterns in the naked DNA sample were similar between dThd and BrdU medium Similar results were obtained when nucleo-some positioning was mapped in the opposite direction from the EcoRV site with probe B (Fig S1, lanes 5–8)
We next analyzed the nucleosome positioning on pRS801-BAR1 Interestingly, the MNase mapping of nucleosomes with probe A showed that the three bands *b to *d (especially band *c) were not protected
in a chromatin sample of pRS801-BAR1 (Fig 2A, lane 9), indicating that insertion of A34disrupts nucleosome positioning more remarkably than that of
T3CCT6CT5GCT5CT7 In BrdU medium, all of the bands that corresponded to the linker regions were eliminated and the bands *a, *b and *e were more clearly detected than in dThd medium (Fig 2B, lanes 9 versus 11) These results were also confirmed by indi-rect end-labeling with probe B (Fig S1, lanes 9–12) Taken together, these observations indicate that nucleosome positioning on pRS001- and pRS801-BAR1 were modestly and almost completely disrupted, respectively, by BrdU
Effect of BrdU on BAR1 expression
We measured the expression of the BAR1 gene on the model plasmids by Northern blot analysis In dThd medium, the BAR1 mRNA level for pRS-BAR1 was
30 times less in the MATa cells than in the MATa cells (Fig 3A) These results indicate that the episomal BAR1gene is regulated similarly to the genomic BAR1 gene In the MATa cells, the mRNA level was 2.4 times higher in BrdU medium than in dThd medium (Fig 3B) Insertion of T3CCT6CT5GCT5CT7 (pRS001-BAR1) slightly increased the mRNA level in dThd medium, but additionally increased it in BrdU medium (Fig 3B) On the other hand, insertion of A34 (pRS801-BAR1) significantly increased the mRNA level in dThd medium and markedly increased it in
Trang 4a b c d
a
b
c
d
2 op
c d
a b
c d
a b
c
e
a b
c
e
a b
pRS-BAR1
C D
C D
pRS001-BAR1
pRS801-BAR1
M
2.0
1.5
1.0
0.5
(kb)
I
II
III
IV
V
A
Lane
1
3
5
7
9
11
Fig 2 Nucleosome positioning on pRS-BAR1 and its derivatives Chromatin (indi-cated by C) and naked DNA (indi(indi-cated by D) samples were prepared from cells
transfect-ed with the plasmid indicattransfect-ed and culturtransfect-ed
in dThd or BrdU medium as indicated The samples were partially digested with
5 UÆmL)1(odd-numbered lanes) or 0.5 UÆmL)1(even-numbered lanes) MNase, completely digested with XhoI, and sub-jected to the indirect end-labeling analysis with probe A as described in Materials and Methods At least three independent analy-ses for each plasmid gave similar results (A) Autoradiography of MNase cleavage pat-terns DNA size markers (M), positions of nucleosomes I–V and a2 operator (a2 op) are shown to the left Specific cleavage sites on naked DNA samples are marked with *a, *b, *c, *d and *e Ellipses with dotted lines indicate nucleosomes whose positioning is unstable Open stars on some lanes denote bands that changed in BrdU medium The transcriptional start site and the KpnI site are indicated by a bent arrow and arrowheads, respectively (B) Densito-metric profiles of autoradiography The odd-numbered lanes in (A) were
densitometrical-ly scanned The vertical dotted lines denote the bands (*a, *b, *c, *d and *e) specific to naked DNA The positions of stable nucleo-somes are shown at the top.
Trang 5BrdU medium (Fig 3B) We also examined the effect
of BrdU on pYBT1 containing the thymidine kinase
(TK) gene driven by the yeast constitutive ADH1
pro-moter as a control plasmid BrdU did not significantly
affect the expression of the gene (Fig 3B) These
results show that the levels of expression of the BAR1
gene are parallel to those of the disruption of
nucleo-some positioning in dThd and BrdU medium
To confirm that our experimental conditions can
induce specific genomic genes, we examined the
expres-sion of some genomic genes having an AT-tract on
their promoter regions The DED1 gene, having
T3CCT6CT5GCT5CT7, and the MAK16 gene, having
T24, were significantly up-regulated by the addition of
BrdU (Fig 4), suggesting that the mechanism found in
the episomal genes also operates in the genomic genes
Effect of BrdU on a2/Mcm1 repressor–operator
complex
We examined whether the a2⁄ Mcm1 repressor changes
its binding to the a2 operator on pRS-BAR1 upon
incorporation of 5-bromouracil by an in vivo UV
photofootprinting assay The a2⁄ Mcm1 operator has
numerous thymine bases necessary for recognition by
the repressor [28] and thus its binding to the repressor
may be disturbed by substitution of 5-bromouracil In
the noncoding strand of a naked DNA sample, several
thymine dimers were found around the a2 operator
(Fig 5A, lane 2), but three sites (marked with + in Fig 5A) were protected in chromatin samples (Fig 5A, lanes 1 and 3) Although slight differences in the thymine dimers formed were observed in the naked DNA samples containing thymine or 5-bromouracil (Fig 5A, lanes 2 versus 5), the three thymine dimers were equally protected in the chromatin samples con-taining thymine or 5-bromouracil (Fig 5A, lanes 4 and 6) Similar results were obtained with the coding strand (Fig 5B) These results show that BrdU does not weaken the formation of the a2⁄ Mcm1 repressor– operator complex, excluding the possibility that the
DED1
MAK16
ACT1
Fig 4 Northern blot analysis of genomic genes Total RNA sam-ples prepared from the MATa cells cultured in dThd or BrdU med-ium were subject to Northern blot analysis with probes derived from the genes indicated.
T B pRS-BAR1 0
2 4 6
T B pRS001-BAR1
##
T B pRS801-BAR1
###
T B
pYBT1
20
0 MAT MATa
pRS-BAR1 10
30
Fig 3 Gene expression profiles of pRS-BAR1, its derivatives and a reference plasmid (A) BAR1 mRNA levels in MATa and MATa cells Total RNA and DNA samples were prepared from the cell type indicated in dThd medium and subjected to Northern and Southern blot analy-ses as described in Materials and Methods BAR1 mRNA levels were expressed relative to actin mRNA levels after normalization by copy numbers of plasmids (B) Effects of BrdU on BAR1 and TK mRNA levels Total RNA and DNA samples were prepared from the MATa cells transfected with the plasmid indicated and cultured in dThd (T) or BrdU medium (B), and processed as in (A) ***P < 0.001 compared with the values in dThd medium.##P < 0.01 and###P < 0.001 compared with the values of pRS-BAR1 in dThd medium Histograms represent means ± standard error At least four independent analyses carried out for each plasmid gave similar results.
Trang 6disruption of nucleosome positioning by BrdU is due
to a decrease in the formation of the a2⁄ Mcm1
com-plex [18]
Effect of BrdU on pRS-BAR1 lacking the promoter activity
We addressed whether the above changes in nucleo-some positioning are affected by the promoter activity
of the BAR1 gene, because transcription factors can affect nucleosome positioning We constructed a plas-mid, pRS-DTA⁄ BAR1, in which the TATA box of BAR1 was disrupted (Fig 6A) [23] With pRS-DTA⁄ BAR1 we were able to determine the change in nucleo-some positioning without the effect of the expression
of BAR1 (Fig 6B)
Disruption of the TATA box caused the disappear-ance of band *d in naked samples of pRS-DTA⁄ BAR1 prepared in dThd and BrdU medium Band *d corre-sponds to the nuclease hypersensitive site located at the TATA box on pRS-BAR1 (Fig 2A, lanes 2 and 4 versus Fig 6C, lanes 2 and 4) as described in the pre-vious reports by Shimizu et al [26] and Cooper et al [23] In dThd medium, the MNase cleavage pattern in
a chromatin sample of pRS-DTA⁄ BAR1 showed well-ordered nucleosome positioning (Fig 6C, lane 1), which was identical to that on the chromatin sample
of pRS-BAR1 (Fig 2A, lane 1), except for the pres-ence of band *d
In BrdU medium, the bands corresponding to the linker regions between nucleosomes I–III were broader, and the two bands (*b and *c) that were pro-tected in the chromatin sample prepared in dThd med-ium became more evident (Fig 6C, lane 3) When
dThd
C D C
BrdU
C D C
+
+
+
+
+
+
+
+ +
C D C
BrdU
C D C
+
+
+
+
+
+
B
1 2 3 4 5 6 7 8 9 10 11 12
Fig 5 In vivo UV photofootprinting of a2 operator on pRS-BAR1.
Intact cells (indicated by C) and naked DNA (indicated by D) were
irradiated with UV at dosages of 250 mJ (lanes 1, 4, 7 and 10),
500 mJ (lanes 3, 6, 9 and 12) and 60 mJ (lanes 2, 5, 8 and 11) UV
photoproducts were analyzed using primer extension mapping on
the noncoding (A) and the coding (B) strands as described in
Mate-rials and Methods The a2 operator sequence is shown to the left
of each panel The thymine bases protected from UV damage are
indicated by +.
a
b c
e I
II III IV
V
a
b c
e
BrdU
CAGTATAAAAGTG
TATA pRS-BAR1
CAGTGGATCCGTG
Bam H I pRS-ΔTA/BAR1
pRS-ΔTA/BAR1 dThd
2 op Lane 1 2 3 4
B
ACT1
BAR1
/BA
R1
pRS-C D
C D
Fig 6 Nucleosome positioning on p0RS-DTA ⁄ BAR1 (A) Sequences of the TATA box and disrupted TATA box of the BAR1 pro-moter (B) Northern blot analysis of the BAR1 gene (C) Autoradiography of MNase cleavage patterns Nucleosome positioning was analyzed as in Fig 2.
Trang 7cultured in dThd or BrdU medium, the overall band
patterns were almost identical in the chromatin
sam-ples between pRS-BAR1 and pRS-DTA⁄ BAR1, except
for the presence of band *d These results indicate that
BrdU changes nucleosome positioning in the absence
of the transcription factors involved In line with this
observation, we have shown that incorporation of
BrdU into DNA converts the DNA structure into an
unusual conformation [18] It is thus reasonable to
suggest that such a structural change in DNA induced
by BrdU affects nucleosome positioning and results in
altered gene expression at particular regions
Discussion
We were able to show a positive correlation between
gene expression and the disruption of nucleosome
posi-tioning in yeast cells harboring minichromosomes and
cultured with BrdU as the only source of thymine To
validate this observation, BrdU must directly affect
DNA structure, but not interactions between DNA
and DNA-binding proteins, to induce a change in
nucleosome positioning In our model plasmids used,
the binding of a2⁄ Mcm1 repressor to its legitimate
binding site has a critical role in the formation of
sta-bly ordered nucleosomes As expected, BrdU did not
significantly affect the formation of the a2⁄ Mcm1
repressor–operator complex [18] In support of this,
expression of the genomic STE2 gene regulated by the
a2⁄ Mcm1 complex was not affected by the addition of
BrdU (data not shown) In addition, some
DNA-bind-ing proteins and enzymes examined to date cannot
functionally distinguish between 5-bromouracil and
thymine on DNA [29–32] These results suggest that
an unusual DNA conformation induced by BrdU is
the primary cause of altered nucleosome positioning
In fact, we have demonstrated that the incorporation
of 5-bromouracil into DNA reduces the bending of
DNA [11] and converts to A-form-like DNA or a rigid
DNA structure [18] However, a possibility cannot be
ruled out that a change in interactions between
5-bro-mouracil-substituted DNA and specific proteins may
additionally affect nucleosome positioning
We showed here that levels of BAR1 expression are
parallel to those of the destabilization of nucleosome
positioning with the use of model plasmids In our
pre-vious study employing different minichromosomes [18],
disruption of nucleosome positioning by BrdU was
shown to depend on the length and sequence specificity
of AT-tracts located at particular sites of
minichromo-somes In this study, destabilization of nucleosome
positioning by BrdU did not require the presence of
the promoter or expression of the BAR1 marker gene
As shown in the Results, AT-tracts alone can destabi-lize nucleosome positioning in dThd medium How-ever, not all AT-tracts have the ability to induce the destabilization of nucleosome positioning or expression
of the marker genes
The nucleosome disruption did not lead to full expression of the BAR1 gene This can be explained by the absence of an activator function of Mcm1 in MATa cells Mcm1 acts as an activator of a-cell-spe-cific genes in MATa cells, whereas it acts as a repressor
in MATa cells Taken together, nucleosome position-ing seems to be sufficient for full repression of genes [27] Also, disruption of it seems to be sufficient for full derepression of genes if necessary transcription fac-tors are supplied
Can our findings obtained with the episomal genes apply to genomic loci? The episomal BAR1 gene was shown to behave similarly to the genomic BAR1 gene [27] when A34was inserted into their promoters Simi-larly, the genomic DED1 gene was significantly up-reg-ulated by BrdU, as the episomal BAR1 gene has the AT-tract derived from the DED1 promoter Further-more, the genomic MAK16 gene having T24on its pro-moter region was also up-regulated by BrdU These results suggest that episomal and genomic genes behave similarly, and prove to be useful in studying gene regulation controlled by the higher-order struc-ture of chromatin
However, the presence of AT-tracts does not always affect the expression of their adjacent genes if the genes have a strong promoter Their promoter regions seem to be reluctant to form a stable nucleosome structure In these genes, BrdU does not seem to cause
an additional change in the nucleosome structure sur-rounding the genes and increase their expression For example, the expression of the TK gene driven by the yeast constitutive ADH1 promoter on pYBT1 was not affected by BrdU (Fig 3B) In the case of TALS-GFP, EGFP expression is driven by the ADH1 promoter (Fig S2) pOM801-GFP, having A34 inserted in the ADH1 promoter region (Fig S2), showed significantly increased promoter activity even in dThd medium (Fig S4) However, BrdU did not further increase the expression of GFP (Fig S4) or the state of the already opened nucleosome structure in pOM801-GFP (Fig S3) These results support our hypothesis that BrdU induces the expression of genes through the dis-ruption of nucleosome positioning on their promoter regions
In mammalian cells, BrdU is thought to induce the expression of genes when AT-tracts are located adja-cent to their promoters, similar to yeast systems [11]
In contrast to yeast genes, most of the mammalian
Trang 8genes are silenced during lifetime, embedded in
con-densed chromatin As described previously, BrdU can
restore the expression of silenced genes [9,10], and
BrdU-responsive genes are frequently located on
inac-tive chromatin regions, such as AT-rich Giemsa-dark
bands of human chromosomes [11,12] In this context,
BrdU seems to disrupt nucleosome positioning around
AT-rich condensed chromatin and results in the
induc-tion of the expression of silenced or repressed genes
Finally, the data of this study may lead to a new
understanding of the molecular mechanism of BrdU
They may answer the new and old question of why
BrdU modulates the expression of particular genes
associated with cellular differentiation and senescence
Materials and methods
Plasmids and yeast strains
To construct BAR1 expression plasmid pRS-BAR1, the
)500 to +2064 sequence containing the promoter with a
KpnI site ()158 to )154) [27], a coding sequence and
300 bp of 3¢-UTR of the BAR1 gene, were cloned into the
XhoI-SacI site of pRS424DKpnI in which one KpnI site was
filled in Plasmid pRS-BAR1 derivatives were constructed
by inserting oligonucleotides into the KpnI site of
pRS-BAR1 to yield pRS001-pRS-BAR1 (T3CCT6CT5GCT5CT7) and
pRS801-BAR1 (A34)
To disrupt a TATA box in pRS-BAR1, the sequence of
the TATA box at )134 (TATAAAA) was changed to a
sequence (TGGATCC) that contained a BamHI site by
amplifying a KpnI-BglII sequence of pRS-BAR1 with the
following two primers: 5¢ -TATTGGTACCGTGTGTTTT
TTGATAACAGTGGATCCGTG-3¢ and 5¢-GTGGAAGA
TCTATGCTCATTATAAGTACTC-3¢ The amplified
sequence was digested with KpnI and BglII, and cloned into
the KpnI–BglII site of pRS-BAR1 to yield pRS-DTA⁄
-BAR1, which lacks a functional TATA box
These plasmids were introduced into the yeast
dThd-autotrophic strains YKH2 (MATa ura2-52 trp1 his3 leu2
cdc21::LEU2 pYBT1) or YKH4 (MATa ura2-52 trp1 his3
leu2 cdc21::LEU2 pYBT1) established as described
previ-ously [2] Plasmid pYBT1 contains the herpes simplex virus
TKgene driven by the yeast ADH1 promoter
Chromatin preparation and nuclease digestion
Yeast cells harboring plasmids were selected in SC medium
(2% glucose, 0.67% yeast nitrogen base without amino
acids) supplemented with appropriate amino acids (except
for tryptophan) and 1 mm dThd Cells were grown in
30 mL of medium containing 1 mm dThd or BrdU at 30C
for 15 h to an optimal density of 0.6–1.0 at 600 nm
Chro-matin and naked DNA samples were prepared according to
the method of Balasubramanian & Morse [33] Each sample was digested with MNase (Takara, Kyoto, Japan) at 37C for 10 min The reactions were initiated by the addition of 0.15% Nonidet P-40, and halted by the addition of SDS and proteinase K Samples were purified with a phe-nol⁄ chloroform extraction and ethanol precipitation
Indirect end-labeling DNA samples were completely digested with XhoI or EcoRV together with RNase A, run on a 1.4% agarose gel and transferred on to a Nylon membrane (Biodyne B, Pall, Port Washington, NY, USA) followed by cross-linking with
UV light (Stratalimker 2400, Stratagene, La Jolla, CA, USA) The membrane was incubated at 65C for 16 h in hybridization solution (0.5 m Na-Pi, 1 mm ETDA and 7% SDS] containing a probe labeled with [a-32P] dCTP using a random-primed DNA labeling kit (Mega-prime, Amersham, Piscataway, NJ, USA) The XhoI–MspI fragment of pRS-BAR1 was used as a probe to detect nucleosomes in sam-ples digested with XhoI Likewise, a sequence amplified with the primers ATCTTATAATTATCGAGATCG-3¢ and 5¢-AAGTGTTCCACTGTCTAGTTTG-3¢ from pRS-BAR1 was used as a probe to detect nucleosomes in samples digested with EcoRV After washing, the membrane was subjected to autora-diography and densitometric analysis using an image analyzer FLA-5000 (FUJIFILM, Tokyo, Japan)
Gene expression analysis Total RNA samples were prepared, subjected to electropho-resis on a 1% formaldehyde agarose gel and blotted on to
a Nylon membrane as described previously [34] To deter-mine copy numbers of plasmids in yeast cells, DNA sam-ples were prepared as described above The DNA samsam-ples were digested with XhoI to linearize and run on a 1% aga-rose gel and blotted on to a Nylon membrane The mem-brane was hybridized with appropriate 32P-labeled BAR1,
TK, DED1, MAK16 or ACT1 coding sequences as probes After washing, the membrane was subjected to autoradiog-raphy followed by imaging analysis as described above
In vivo UV photofootprinting
An in vivo UV photofootprinting assay was performed as described previously [35,36] Yeast cells harboring a plasmid were grown in SC (Trp)) medium containing 1 mm dThd or BrdU at 30C for 15 h, irradiated with UV at 254 nm using
a Stratalinker 2400, and DNA samples purified As controls, DNA samples were purified from unirradiated cells and irra-diated similarly Sites and levels of the formation of UV photoproducts were determined by primer extension map-ping, using the IR-Dye-800-labeled MS-9 primer ()387 to )358 of the BAR1 coding strand) and MS-10 primer ()87 to
Trang 9)116 of the BAR1 noncoding strand) with a LI-COR 4000L
DNA sequencer (LI-COR, Lincoln, NE, USA) [26,37]
Acknowledgements
This work was supported in part by Grants-in-Aid for
Scientific Research from the Ministry of Education,
Science and Culture of Japan
References
1 Michishita E, Nakabayashi K, Suzuki T, Kaul SC,
Ogino H, Fujii M, Mitsui Y & Ayusawa D (1999)
5-Bromodeoxyuridine induces senescence-like phenomena
in mammalian cells regardless of cell type or species
J Biochem (Tokyo) 126, 1052–1059
2 Fujii M, Ito H, Hasegawa T, Suzuki T, Adachi N &
Ayusawa D (2002) 5-Bromo-2¢-deoxyuridine efficiently
suppresses division potential of the yeast Saccharomyces
cerevisiae Biosci Biotechnol Biochem 66, 906–909
3 Wilt FH & Anderson M (1972) The action of
5-bromo-deoxyuridine on differentiation Dev Biol 28, 443–447
4 Weintraub H (1973) Size of the BUdR sensitive targets
for differentiation Nat New Biol 244, 142–143
5 Goz B (1977) The effects of incorporation of
5-haloge-nated deoxyuridines into the DNA of eukaryotic cells
Pharmacol Rev 29, 249–272
6 Suzuki T, Minagawa S, Michishita E, Ogino H,
Fujii M, Mitsui Y & Ayusawa D (2001) Induction of
senescence-associated genes by 5-bromodeoxyuridine in
HeLa cells Exp Gerontol 36, 465–474
7 Minagawa S, Nakabayashi K, Fujii M, Scherer SW and
Ayusawa D (2005) Early BrdU-responsive genes
consti-tute a novel class of senescence-associated genes in
human cells Exp Cell Res 304, 552–558
8 Zakharov AF, Baranovskaya LI, Ibraimov AI,
Benjusch VA, Demintseva VS & Oblapenko NG (1974)
Differential spiralization along mammalian mitotic
chromosomes II 5-bromodeoxyuridine and
5-bro-modeoxycytidine-revealed differentiation in human
chromosomes Chromosoma 44, 343–359
9 Suzuki T, Yaginuma M, Oishi T, Michishita E,
Ogino H, Fujii M & Ayusawa D (2001)
5-Bromodeoxy-uridine suppresses position effect variegation of
transgenes in HeLa cells Exp Cell Res 266, 53–63
10 Fan J, Kodama E, Koh Y, Nakao M & Matsuoka M
(2005) Halogenated thymidine analogues restore the
expression of silenced genes without demethylation
Cancer Res 65, 6927–6933
11 Suzuki T, Michishita E, Ogino H, Fujii M &
Ayusawa D (2002) Synergistic induction of the
senescence-associated genes by 5-bromodeoxyuridine
and AT-binding ligands in HeLa cells Exp Cell Res
276, 174–184
12 Minagawa S, Nakabayashi K, Fujii M, Scherer SW & Ayusawa D (2004) Functional and chromosomal clus-tering of genes responsive to 5-bromodeoxyuridine in human cells Exp Gerontol 39, 1069–1078
13 Satou W, Suzuki T, Noguchi T, Ogino H, Fujii M & Ayusawa D (2004) AT-hook proteins stimulate induc-tion of senescence markers triggered by 5-bromodeoxy-uridine in mammalian cells Exp Gerontol 39, 173–179
14 Kingston RE, Bunker CA & Imbalzano AN (1996) Repression and activation by multiprotein complexes that alter chromatin structure Genes Dev 10, 905–920
15 Wolffe AP & Hayes JJ (1999) Chromatin disruption and modification Nucleic Acids Res 27, 711–720
16 Morse RH (2007) Transcription factor access to pro-moter elements J Cell Biochem 102, 560–570
17 Shimizu M, Mori T, Sakurai T & Shindo H (2000) Destabilization of nucleosomes by an unusual DNA conformation adopted by poly(dA)⁄ poly(dT) tracts in vivo EMBO J 19, 3358–3365
18 Miki K, Shimizu M, Fujii M, Hossain MN &
Ayusawa D (2008) 5-Bromouracil disrupts nucleosome positioning by inducing A-form-like DNA conforma-tion in yeast cells Biochem Biophys Res Commun 368, 662–669
19 Lin S, Lin D & Riggs AD (1976) Histones bind more tightly to bromodeoxyuridine-substituted DNA than to normal DNA Nucleic Acids Res 3, 2183–2191
20 Fasy TM, Cullen BR, Luk D & Bick MD (1980) Stud-ies on the enhanced interaction of halodeoxyuridine-substituted DNAs with H1 histones and other polypep-tides J Biol Chem 255, 1380–1387
21 Roth SY, Shimizu M, Johnson L, Grunstein M & Simpson RT (1992) Stable nucleosome positioning and complete repression by the yeast alpha 2 repressor are disrupted by amino-terminal mutations in histone H4 Genes Dev 6, 411–425
22 Ganter B, Tan S & Richmond TJ (1993) Genomic foot-printing of the promoter regions of STE2 and STE3 genes in the yeast Saccharomyces cerevisiae J Mol Biol
234, 975–987
23 Cooper JP, Roth SY & Simpson RT (1994) The global transcriptional regulators, SSN6 and TUP1, play dis-tinct roles in the establishment of a repressive chroma-tin structure Genes Dev 8, 1400–1410
24 Davie JK, Trumbly RJ & Dent SY (2002) Histone-dependent association of Tup1-Ssn6 with repressed genes in vivo Mol Cell Biol 22, 693–703
25 Ruiz C, Escribano V, Morgado E, Molina M & Mazon MJ (2003) Cell-type-dependent repression of yeast a-specific genes requires Itc1p, a subunit of the Isw2p-Itc1p chromatin remodelling complex
Microbiology 149, 341–351
26 Shimizu M, Roth SY, Szent-Gyorgyi C & Simpson RT (1991) Nucleosomes are positioned with base pair
Trang 10precision adjacent to the alpha 2 operator in
Saccharomyces cerevisiae EMBO J 10, 3033–3041
27 Morohashi N, Yamamoto Y, Kuwana S, Morita W,
Shindo H, Mitchell AP & Shimizu M (2006) Effect of
sequence-directed nucleosome disruption on
cell-type-specific repression by alpha2⁄ Mcm1 in the yeast
gen-ome Eukaryot Cell 5, 1925–1933
28 Zhong H, McCord R & Vershon AK (1999)
Identifica-tion of target sites of the alpha2-Mcm1 repressor
complex in the yeast genome Genome Res 9, 1040–
1047
29 Goeddel DV, Yansura DG, Winston C &
Caruthers MH (1978) Studies on gene control regions
VII Effect of 5-bromouracil-substituted lac operators
on the lac operator–lac repressor interaction J Mol Biol
123, 661–687
30 Brennan CA, Van Cleve MD & Gumport RI (1986)
The effects of base analogue substitutions on the
cleav-age by the EcoRI restriction endonuclease of
octade-oxyribonucleotides containing modified EcoRI
recognition sequences J Biol Chem 261, 7270–7278
31 Hayakawa T, Ono A & Ueda T (1988) Synthesis of
decadeoxyribonucleotides containing 5-modified uracils
and their interactions with restriction endonucleases Bgl
II, Sau 3AI and Mbo I (nucleosides and nucleotides
82) Nucleic Acids Res 16, 4761–4776
32 Risse G, Jooss K, Neuberg M, Bruller HJ & Muller R
(1989) Asymmetrical recognition of the palindromic
AP1 binding site (TRE) by Fos protein complexes
EMBO J 8, 3825–3832
33 Balasubramanian B & Morse RH (1999) Binding of
Gal4p and bicoid to nucleosomal sites in yeast in
the absence of replication Mol Cell Biol 19, 2977–
2985
34 Takayama S, Fujii M, Kurosawa A, Adachi N &
Ayus-awa D (2007) Overexpression of HAM1 gene detoxifies
5-bromodeoxyuridine in the yeast Saccharomyces cerevi-siae Curr Genet 52, 203–211
35 Murphy MR, Shimizu M, Roth SY, Dranginis AM & Simpson RT (1993) DNA-protein interactions at the
S cerevisiaealpha 2 operator in vivo Nucleic Acids Res
21, 3295–3300
36 Shimizu M & Mitchell AP (2003) Hap1p photofoot-printing as an in vivo assay of repression mechanism in Saccharomyces cerevisiae Methods Enzymol 370, 479– 487
37 Morohashi N, Nakajima K, Kuwana S, Tachiwana H, Kurumizaka H & Shimizu M (2008) In vivo and in vi-tro footprinting of nucleosomes and transcriptional acti-vators using an infrared-fluorescence DNA sequencer Biol Pharm Bull 31, 187–192
Supporting information
The following supplementary material is available: Fig S1 Nucleosome positioning on pRS-BAR1 and its derivatives
Fig S2 Schematic representation of TALS-GFP and its derivative
Fig S3 Nucleosome positioning on TALS-GFP and pOM801-GFP
Fig S4 Effects of BrdU on EGFP mRNA levels This supplementary material can be found in the online version of this article
Please note: As a service to our authors and readers, this journal provides supporting information supplied
by the authors Such materials are peer-reviewed and may be re-organized for online delivery, but are not copy-edited or typeset Technical support issues arising from supporting information (other than missing files) should be addressed to the authors