| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Laboratory of Molecular Genetics, National Institute on Aging, National Institutes of Health, Baltimore, Maryland 21224, USA
1Correspondence: Laboratory of Molecular Genetics, National Institutes on Aging, NIH, 5600 Nathan Shock Dr., Baltimore, MD 21224, USA. E-mail: vbohr{at}nih.gov
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Key Words: life span • aging • population doubling • rDNA
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
It has been speculated that aging is accompanied by changes at the rDNA
locus. The proximity of multiple rDNA genes to one another makes them
potentially susceptible to homologous recombination resulting in
possible gain or loss of rDNA gene copies. The rDNA copy number has
been the focus of a number of studies, many of which indicate that
there is a selective loss of rDNA during the course of aging. An
age-related loss of rDNA has been reported in postmitotic brain tissues
of beagle dogs (1)
, human myocardium and cerebral cortex
(2
, 3)
, and mouse brain, spleen, and kidney tissues
(4)
. Recent studies have implicated changes in rDNA in the
cellular senescence of Saccharomyces cerevisiae (reviewed in
ref 5
). Extrachromosomal rDNA circles (ERCs) accumulate in
budding yeast during cellular senescence and the introduction of
exogenous ERCs accelerates yeast aging (6)
. Studies of
sgs1 deletion mutants in yeast show increased recombination
at the rDNA locus (7
, 8)
, an accelerated accumulation of
ERCs (6)
, and a significant reduction in the average life
span (9)
. Thus, changes in the rDNA copy number have been
associated with cellular senescence in yeast.
The in vitro senescence of primary fibroblasts has been used
as a model system for mammalian aging studies. A correlation exists
between the life span of a species and the number of times its cells
can divide in culture (10
, 11)
. Also, several studies
report that the replicative capacity of cells in culture generally
decreases with increasing age of the donor (10
, 11)
.
However, this view has been challenged by recent studies
(12)
that find no correlation between donor age and
in vitro life span. Nonetheless, a connection between
cellular senescence and organismal aging is supported by the premature
aging phenotype of patients with Werner syndrome (WS). Patients with WS
have an early onset of characteristics of aging such as skin atrophy,
early graying and loss of hair, and increased incidence of a number of
age-related diseases such as malignant neoplasms, atherosclerosis, type
II diabetes mellitus, osteoporosis, and cataracts
(13
14
15)
. Primary fibroblasts established from WS patients
divide fewer times in culture than cells derived from normal
individuals (16
, 17)
. Notably, the gene that is defective
in WS (known as WRN) is a human homologue of the
aforementioned yeast SGS1 gene. Both WRN and Sgs1 are
members of a family of RecQ-like DNA helicases (7
, 8
, 18)
and have DNA helicase activity in vitro (19
, 20)
. Although the exact function of WRN in humans is unknown,
the hyper-recombination of rDNA genes in sgs1 mutants
suggests that the lack of a functional WRN protein in WS cells may
result in hyper-recombination at the rDNA loci. That mutations in the
yeast SGS1 and human WRN genes can cause
premature aging phenotypes suggests that there may be a common
conserved mechanism of aging involving changes in rDNA in both yeast
and mammals.
Other changes at the rDNA loci may be relevant to organismal and
cellular aging. The level of expression of rRNA genes could influence
ribosomal function and protein synthesis. Base modifications (such as
methylation) within rDNA sequences may affect the expression of rRNA
genes. The extent of methylation in promoters of RNA polymerase II (pol
II) -transcribed genes has been shown to influence transcription
(21
22
23)
. Some studies have shown increases in methylation
in promoters of specific pol II-transcribed genes during in
vivo and in vitro aging (24
, 25)
, despite
the finding that overall genome methylation levels decrease both with
age in animal tissues and during cellular senescence (26
, 27)
. Although an increase in rDNA methylation in aging mouse
tissues has been reported (28)
, the methylation of rDNA
genes during in vivo or in vitro aging has not
been examined in detail. It is also unclear whether methylation of rRNA
genes influences RNA polymerase I (pol I) transcription.
Our main aim was to study the role of rDNA in premature and normal human aging by looking for changes in rDNA genes during cellular senescence. We have examined the rDNA gene copy number, expression level, and methylation during the in vitro senescence of primary fibroblasts derived from both normal individuals (young and old donors) and WS patients. We find that both the copy number of rDNA genes and the steady-state levels of 28S rRNA remain constant throughout the in vitro life span of both normal and WS fibroblasts. However, senescent normal and WS fibroblasts show significant increases in the methylation of cytosine in rDNA genes. Notably, this increase in methylation is most pronounced in the abbreviated life span of WS fibroblasts.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
. All cultures were grown in minimal essential medium with Earle’s
salts (Gibco BRL, Grand Island, N.Y.) supplemented with essential amino
acids (0.25 µg/ml), nonessential amino acids (0.25 µg/ml), vitamins
(1 mg/ml) (Gibco BRL), 15% fetal bovine serum (Gibco BRL), penicillin
(100 U/ml), and streptomycin (100 µg/ml) (Gibco BRL). Cultures were
maintained at 37°C in a humidified atmosphere containing 5%
CO2 and had viabilities of greater than 90%, as
judged by the trypan dye exclusion assay.
|
Isolation of genomic DNA
Viable cells (1–5x106 cells) were washed
twice with phosphate-buffered saline (PBS), suspended in lysis solution
(0.5 M Tris [pH 8.0], 1% sodium dodecyl sulfate, 0.5 mg/ml
proteinase K) and incubated at 37°C for 16–24 h. Next, saturated
NaCl (1/4 volume) was mixed vigorously with the lysate, and the
samples were centrifuged at 2500 rpm for 30 min. DNA in the supernatant
was precipitated with ethanol and the precipitate was collected by
centrifugation, washed with 70% ethanol, and resuspended in 1 ml of TE
(10 mM Tris-HCl [pH 8.0],1 mM EDTA). The DNA was subsequently treated
with 100 µg of RNase A for 3 h at 37°C, reprecipitated (with
0.2 volumes of 11 M ammonium acetate and 2–2.5 volumes of ethanol),
and washed with 70% ethanol. The DNA was finally dissolved in 100–200
µl TE and concentration determined by UV spectrophotometry. The
relative purity of each DNA sample was ensured by examining the 260/280
nm absorption ratio. For determining the rDNA copy number, 2 µg of
DNA was incubated with EcoRI (10 units) in the appropriate
restriction buffer for 16–20 h.
Analysis of methylation by 5-methylcytosine-sensitive restriction
enzymes
The restriction endonucleases HpaII and
MspI were used to study sequence-specific methylation
patterns by Southern blot analysis as described previously (28
, 29)
. HpaII will cut the sequence CCGG but is
inhibited if the internal cytosine is methylated, whereas
MspI cuts both CCGG and CmeCGG
efficiently (30)
. In parallel, equivalent amounts (1 µg)
of genomic DNA were incubated with either HpaII or
MspI in buffers specified by the manufacturer for 16–20 h
at concentrations of 4–5 units/µg of DNA. Southern blotting
techniques (described below) were used to compare
HpaII-digested DNA side-by-side with
MspI-digested DNA for changes in methylation of CpG sites in
CCGG sequences.
Southern blotting and hybridizations
After restriction endonuclease digestion, DNA samples were
electrophoresed on agarose gels (0.5, 0.8, or 1%). After
electrophoresis, the gels were soaked in 0.25 M HCl for 30 min, then in
denaturing solution (1.5 M NaCl, 0.5 N NaOH) for 45 min, and finally in
neutralizing solution (1 M Tris [pH 7.4], 1.5 M NaCl) for 45 min. The
DNA was transferred to nylon membranes (Hybond N; Amersham, Little
Chalfon, U.K.) using a posiblot apparatus (Stratagene, San Diego,
Calif.) with 10 x SSPE (1.5 M NaCl, 0.1 M
NaH2PO4, 0.01 M EDTA) as
the transfer buffer (31)
. After UV cross-linking of DNA to
the membrane, the blots were incubated in Hybrisol I (Oncor) for
12 h at 45°C, followed by hybridization for 12–14 h in the same
solution supplemented with 
107 dpm
denatured probe/ml. After hybridization, the membranes were washed
1) twice with 2 x SSPE, 0.1% sodium dodecyl sulfate
(SDS) for 15 min at 25°C, 2) once with 0.1 x SSPE,
0.1% SDS for 15 min at 25°C, and 3) twice with 0.1 x SSPE, 0.1% SDS for 30 min at 60°C. The membranes were rinsed in
2 x SSPE and air dried. Radioactivity associated with specific
28S rDNA and dihydrofolate reductase (DHFR) bands was measured using a
PhosphorImager (Molecular Dynamics, Sunnyvale, Calif.) and quantitated
with ImageQuant software.
Preparation and analysis of RNA
Metabolically active cells (1–3x106)
were washed twice in PBS, scraped from the petri dish in 1 ml STAT-60
(Teltest-B), and subsequently frozen at -70°C. After thawing, 0.3 ml
chloroform was added, the sample was mixed vigorously for 15 s and
the organic and aqueous phases were separated by centrifugation at
14,000 g for 20 min at 4°C. The aqueous phase was
collected and RNA was precipitated in 0.7 ml isopropanol at -20°C
overnight. The precipitate was collected by centrifugation at 14,000
g for 10 min at 4°C, rinsed briefly in 75% ice-cold
ethanol, air dried, and finally resuspended in 30–100 µl of sterile
DEPC-treated water. The concentration of the RNA samples was measured
by UV spectrophotometry. RNA (1 µg) was electrophoresed on denaturing
gel systems using formaldehyde and formamide as described previously
(32)
. The gels were scanned using a fluoroimager
(Molecular Dynamics) to measure ethidium bromide fluorescence and the
amount of rRNA (28S) in each lane was determined.
DNA probes
A 1.8 kb DHFR genomic DNA probe was isolated from plasmid
pBH31R1.8 (obtained from Dr. Giuseppe Attardi, California Institute of
Technology) as previously described (33)
. The 6.7 kb rDNA
probe from rat recognizing a 7.1 kb EcoRI fragment in the
human sequence (Fig. 1A
) was from the plasmid pEE6.7, described elsewhere
(34)
. Probes were labeled to a high specific activity
(>1.9x109 dpm/µg) with
[
-32P]-dCTP using a random prime labeling
kit (Rediprime II, Amersham Pharmacia Biotech) as specified by the
manufacturer.
|
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
. The WS designations of AG00780 and AG03141 have
been confirmed by genetic analysis (35)
and Western
blotting (36)
, respectively. Cells from each individual
fibroblast culture were counted at near confluence and subcultured
twofold, with each successive split constituting one population
doubling (PDL). This procedure was repeated until the growth of each
culture was negligible. As the growth rate of each culture slowed,
increasing numbers of cells exhibited morphological changes typical of
senescent cells (37)
including enlarged size and greater
numbers of cytoplasmic microfilaments. In ‘senescent’ cultures,
incorporation of bromodeoxyuridine (BrdU) into DNA was undetectable (by
immunofluorescence using an anti-BrdU antibody) in the vast majority of
cells, confirming their lack of replicative potential (data not shown).
Our cell culture protocol allowed the collection, examination, and
comparison of cells and biological (DNA and RNA) samples from normal
and WS fibroblasts at successive points during the entire replicative
life span of each individual culture. At selected passages, total
genomic DNA and RNA were isolated from normal and WS cultures, and
purified DNA samples were analyzed for changes in copy number and
methylation state at the rDNA locus. RNA samples from early and
senescent, normal and WS fibroblasts were compared to determine if
there were changes in the steady-state levels of rRNA.
Primary fibroblasts established from WS patients grew very slowly and
senesced after relatively fewer PDLs than fibroblasts from normal
individuals. Both WS fibroblast cultures had an extremely slow growth
rate after 15 PDLs and negligible growth at 26–27 PDLs. This
is in agreement with the previously reported premature replicative
senescence observed in cells from WS patients (16
, 17)
.
The growth characteristics of normal fibroblast cultures in our study
showed no direct correlation between donor age and in vitro
life span. This is exemplified by the relatively short replicative life
span of a GM8447 (2-day-old) and the long life span of an AG05247
(87-year-old) donor (Table 1)
. These findings support a recent study by
Cristofalo et al. (12)
.
The rDNA copy number does not change significantly with PDLs
First, we analyzed whether in vitro senescence in
human primary fibroblasts from normal and WS donors is accompanied by
changes in the rDNA copy number. We selected four primary fibroblast
cultures from normal individuals (Table 1)
. GM8429 and GM8447 were from
young (1- and 2-day-old, respectively) donors and AG11744 and AG05247
were derived from old (84- and 87-year-old, respectively) donors. Two
primary fibroblast cultures, AG03141 and AG00780, from patients with WS
were also examined (Table 1)
. Genomic DNA was isolated from normal and
WS primary fibroblasts at selected PDLs and restricted with
EcoRI. Southern blot analysis was carried out using equal
amounts of DNA from serial passages of each set of fibroblasts. A probe
for rDNA sequences including the 5.8S rDNA gene, an internal
transcribed spacer region, and most of the 28S rDNA gene (Fig. 1A
) detected a 7.1 kb EcoRI fragment in each DNA
sample. Representative Southern blots from four sets of fibroblasts are
shown in Fig. 1B
. The same blots were cohybridized with a
probe to the single copy housekeeping gene, DHFR. The hybridization
intensity of the DHFR fragment in the same lane was used to correct for
differences in loading of DNA samples. The corrected rDNA hybridization
intensities at selected PDLs of duplicate sets of each fibroblast
culture are presented in Table 2
. Due to differences in hybridization conditions, specific activities of
probes, and exposure times, the numerical values for rDNA hybridization
intensity between individual experiments on different blots (i.e.,
Expt. 1 and Expt. 2 for each fibroblast culture) are not directly
comparable. The mean rDNA hybridization signal for each set of
fibroblast cultures was determined (Table 2)
and the variance from this
mean (normalized to unity) at successive PDLs for each replicate series
of fibroblasts is plotted in Fig. 2
. Despite some experimental variation, this data clearly shows no
systematic increase or decrease in the rDNA copy number during the
in vitro life span of these fibroblast cultures. Notably,
both WS fibroblast cultures maintained the same relative rDNA copy
number throughout their in vitro life span. A similar
analysis of rDNA copy number in fibroblasts from GM8447 (2-day-old
donor) and AG11744 (84-year-old donor) also showed no significant
change in rDNA copy number during cellular senescence (data not shown).
Hence, there are no significant changes in rDNA copy number over the
course of in vitro senescence of any normal (young and old
donors) or WS fibroblasts examined in this study.
|
|
We observed some slower migrating bands that hybridized with our rDNA
probe in a few samples (Fig. 1C
). These additional bands
could easily be discerned in genomic DNA isolated from the most
senescent passages of cultures from old individuals (84- and 87-year
old donors) and WS patients. In these cultures these rDNA species could
also be detected (albeit much less prominently) in the DNA purified
from the passages leading up to senescence. It is important to note
that we could not detect any change in the restriction pattern of the
DHFR gene regardless of passage (data not shown). This suggests that
regions outside of the rDNA locus are efficiently cleaved by
EcoRI even in genomic DNA from the senescent passages, thus
diminishing the possibility of nonspecific inhibition of restriction
occurring in the DNA samples isolated from senescent cultures. These
slower migrating species could either be longer, linear rDNA fragments
or nonlinear rDNA structures, perhaps similar to ERCs in yeast or
blocked replication/recombination intermediates. To determine the
nature of these slow migrating species, DNA samples from the senescent
cultures of normal old donor (AG05247) and WS patients (AG03141 and
AG00780) were electrophoresed through 0.5% and 0.8% agarose gels,
along with linear and superhelical DNA markers. This type of
electrophoretic analysis has been used before to differentiate between
linear and nonlinear DNA molecules (38)
. Linear DNA
molecules migrate to the same position relative to linear DNA markers
in gels regardless of agarose concentration. In contrast, nonlinear DNA
migrates differentially with respect to linear DNA markers in gels
containing different concentrations of agarose. Figure 3
shows the result of this electrophoretic analysis on the DNA
sample from the senescent WS fibroblast culture (AG03141). Note that
the supercoiled circular DNA ladder migrated differentially with
respect to the linear DNA markers in the 0.5% and 0.8% agarose gels.
In both gels, all rDNA-hybridizing bands migrated similarly with
respect to the linear markers. This electrophoretic behavior indicates
that these slower migrating rDNA species are longer linear fragments
and not circular or otherwise nonlinear in nature. The sizes of the
longer, linear rDNA hybridizing fragments are 13 kb and greater than
23 kb.
|
The methylation of rRNA genes increases with cumulative population
doublings
The detection of longer linear fragments in the Southern blot
analysis of genomic DNA from near senescent cells could be due to
inhibition of restriction endonuclease (EcoRI) activity.
Examination of the restriction map of the human rDNA nucleotide
sequence (Fig. 1A
) suggested that the two additional bands
could arise if one or both of the EcoRI site(s) surrounding
the 7.1 kb fragment remained uncut. Inhibition of restriction could be
the result of modifications (such as methylation of adenine and
cytosine) at or near the EcoRI recognition sequence. In
fact, EcoRI has been shown to be particularly sensitive to
cytosine methylation at CpG sites within its recognition sequence.
(30
, 39
, 40)
. We thus investigated this possibility by
measuring the amount of methylation in rDNA genes, using isoschizomeric
restriction enzyme analysis to detect CpG methylation
(28)
.
The restriction endonucleases HpaII and MspI have
the same recognition sequence CCGG. However, methylation of the
internal cytosine in this sequence inhibits HpaII
endonuclease activity but has no effect on MspI activity
(30)
. Thus, these enzymes were used to compare the level
of methylation in rDNA isolated from normal and WS primary fibroblasts
at early and late stages of growth in culture. Equal amounts of genomic
DNA were digested with either HpaII or MspI and
analyzed by Southern blot hybridizations using the rDNA probe described
earlier (Fig. 4
). Treatment with either restriction enzyme resulted in detection of
three restriction fragments of less than 0.7 kb. In addition, when DNA
was digested with HpaII, a broad distribution of higher
molecular weight DNA fragments was observed (Fig. 4)
, indicating
inhibition of restriction due to significant methylation of CpG sites
in rDNA. To estimate the degree of methylation, the hybridization
signal in the smallest restriction fragment of the HpaII
digest was compared to the MspI digest. The signal intensity
of this fragment was consistently less in the HpaII-treated
sample as compared to the MspI-treated sample. This decrease
in the amount of the small fragment also indicates that the restriction
by HpaII is inhibited by methylation of DNA. For the purpose
of comparison, we defined the term ‘methylation index’ as the
MspI/HpaII ratio of the hybridization signal
intensity of the smallest fragment. Increases in this ratio reflect
increased methylation of CpG sites in CCGG sequences between the early
and senescent passages for each cell line (Fig. 5A
). The increase in methylation index during in
vitro life span (Fig. 5B
) ranged from 8% in
DNA isolated from the primary fibroblasts established from young donors
(GM8429, GM8447) to 14–21% in DNA isolated from fibroblasts from old
donors (AG05247, AG11744). The maximal increase in the methylation
index (22–28%) was observed between the early and late passages of WS
fibroblasts (AG03141, AG00780). Thus, in senescent cells, there is a
pronounced increase in the methylation of 28S rDNA. This increase is
maximal in WS fibroblasts even though they go through fewer PDLs than
normal fibroblasts, suggesting that WS cells have an accelerated rate
of methylation. Our isoschizomeric restriction enzyme analysis is
specific for CCGG sequences and does not measure CpG methylation in any
other sequence context. Hence, the increase in methylation measured by
this method is likely an underestimate of the actual increase in
methylation. Thus, there may be an even more dramatic difference in the
increase in methylation associated with senescence in WS cells as
compared with normal cells.
|
|
The steady-state level of 28S rRNA does not change significantly
with cellular senescence
We also examined the expression of rRNA with in vitro
senescence. Equal amounts of total cellular RNA from the early and
senescent passage for each fibroblast culture were electrophoresed on
agarose gels and the amount of 28S rRNA was determined as described in
Materials and Methods. Our results indicate that there was no
significant change in the 28S rRNA/total RNA ratio between the early
and late passage for each individual fibroblast culture (Fig. 6
). Thus, the relative proportion of rRNA appears to be constant over the
in vitro life span of both the normal and WS fibroblasts.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
, 17)
demonstrating that WS cultures generally exhibit
a significantly shorter life span than normal fibroblasts. However, in
our normal fibroblasts, no correlation was observed between donor age
and in vitro life span, a view that has gained more support
recently (12)
. We also find that there is no major change
in rDNA copy number or expression with in vitro aging, but
that there is a marked increase in methylation of rDNA in senescent
cells. This increase is particularly notable in WS cells.
Earlier studies examining the rDNA copy number in yeast and mammalian
tissues have shown an age-related change in the rDNA copy number
(1
2
3
4
, 6)
. In contrast, we could not observe any
significant gain or loss of rDNA units in the normal primary
fibroblasts over their full in vitro life span. Our results
agree with other studies (41
, 42)
showing that in
vitro senescence in cultured mammalian cells is not accompanied by
significant changes in rDNA copy number. The two WS cell cultures also
did not show any change in rDNA copy number with in vitro
senescence. These results suggest that change in rDNA copy number is
probably not associated with or responsible for normal or premature
replicative senescence in mammalian cells in culture. In mammals,
potential age-related changes in the rDNA copy number may be tissue
specific and probably are not associated with the growth of cells in
culture.
Although the rDNA copy number did not change during cellular senescence
in either normal or WS fibroblasts, slower migrating rDNA species could
be detected in senescent passages of normal (old) and WS fibroblasts.
The slower migrating bands could represent aberrant DNA structures or
longer linear fragments that arise due to deficiencies in WRN and
possibly other proteins in senescent cultures. This was an attractive
concept because of the observed accumulation of ERCs during the aging
of yeast cells (6)
, the accelerated accumulation of ERCs
in sgs1 (a homologue of WRN) mutants
(6)
, the putative roles of WRN homologs Sgs1 and FFA-1 in
recombination and replication (7
, 8
, 43)
, and the
association of Sgs1 and WRN with the nucleolus (44
, 45)
.
In yeast, the accumulation of ERCs is shown to cause accelerated
in vitro senescence (6)
. Although the slower
migrating rDNA species (only in the senescent passages of fibroblasts
from normal old donors and WS patients) were suggestive of ERCs or
perhaps stalled replication/recombination intermediates, our
electrophoretic analysis indicates that these are longer linear
fragments. Although our analysis does not rule out the existence of
circular rDNA species in mammalian cells, both the lack of change in
copy number and our inability to detect circular/aberrant rDNA
structures suggest that the rDNA structure-related mechanism that
causes yeast aging may not operate during the in vitro
senescence of cultured normal or WS human cells.
The linear nature of these slower migrating, high molecular weight rDNA
species suggests that the EcoRI restriction site(s) giving
rise to the 7.1 kb hybridizing band might be modified. In support of
this hypothesis, EcoRI is inhibited by methylation in one or
both strands within its recognition sequence (30
, 39
, 40)
and likely by methylation of cytosine or adenine bases nearby
(30
, 46)
. In mammalian cells, CpG sites are the prime
targets for methylation of cytosine. The nucleotide sequence of human
rDNA (GenBank accession number U13369) shows the presence of CpG sites
within and near the relevant EcoRI recognition sequences.
This raises the possibility that methylation at or near
EcoRI sites could be responsible for inhibition of
EcoRI cleavage and the generation of longer, linear rDNA
species in senescent normal old and WS fibroblasts. Our experiments
using isoshizomeric restriction enzyme analysis to digest genomic DNA
indicate that methylation of cytosine residues in 28S rDNA increases
significantly between the first and last passages of each primary
fibroblast culture. The maximum difference in methylation of rDNA
between the early and senescent passages was observed in WS
fibroblasts, whereas rDNA analyzed from fibroblasts of normal old and
young donors showed lower and marginal increases, respectively. In
senescent normal (old) and WS cells, the high degree of methylation of
rDNA correlates precisely with the detection of longer, linear rDNA
fragments after EcoRI restriction. These findings strongly
suggest that the longer, linear rDNA fragments observed in senescent
normal (old) and WS fibroblasts are due to inhibition of
EcoRI cleavage by methylation. Previous studies have shown
increased methylation of the 5' end of the rDNA repeat unit during
aging in mice (28)
. Our results suggest that increased
methylation of rDNA genes occurs during aging in vitro as
well. Methylation of CpG sequences in rDNA appears to be particularly
accelerated in senescent WS fibroblasts. However, the possibility that
rDNA methylation is intrinsically connected with the loss of
replicative potential in normal or WS cells remains to be established.
By promoting the assembly of transcriptionally active DNA into
compacted (inactive) chromatin, increased methylation of cytosine
residues at CpG sites has been correlated with decreased expression of
pol II-transcribed genes (23
, 47)
. Thus, changes in the
methylation state of rDNA could also potentially influence rDNA
transcription and, in turn, ribosomal function and protein synthesis.
However, our observed increases in methylation in the 28S region of
rDNA did not result in any significant change in the steady-state level
of 28S rRNA between the early and senescent passages of any of the
fibroblast cultures used in this study. In agreement, other studies
show that transcription of endogenously methylated sperm rDNA from frog
oocytes or in vitro methylated rDNA gene plasmids occurs at
similar levels as unmethylated rDNA genes (48
, 49)
.
Although our experiments did not analyze the methylation changes
occurring in the promoter region of the rRNA genes, it is unclear from
earlier studies (28
, 50)
whether the level of methylation
even in the rDNA promoter region affects the transcription of rRNA
genes. Further studies are needed both to pinpoint precise sequences in
the rDNA locus that are subject to enhanced methylation during
senescence and to determine how methylation at specific sites affects
rDNA expression.
Methylation of cytosine in rDNA genes could have other potentially
serious consequences. 5-Methylcytosine residues in DNA are deaminated
at a significantly higher rate than unmethylated cytosine
(51)
. Deamination of 5-methylcytosine generates a G:T
mismatch that can potentially result in a G:C to A:T transition
mutation (52)
. Approximately one-third of the point
mutations responsible for human genetic diseases are C
T changes at
CpG sites (53)
. Thus, increased methylation during the
course of in vitro senescence could cause an increased rate
and accumulation of mutations in rDNA genes. Transcription of mutated
rDNA genes could lead to nonfunctional rRNAs, particularly in light of
the extreme conservation of these sequences in evolution. Thus, in this
insidious manner, increased methylation of rDNA genes with age could
result in altered ribosomal function and decreased protein synthesis.
Overall protein synthesis rates have been shown to decrease (40–70%
in mammals) with aging in vivo and in vitro
(54
55
56)
. However, the levels of rRNA do not appear to
decrease with aging (56
; this study). Although the exact
causes remain unclear, lowered protein synthesis rates during aging
have been generally attributed to lowered efficiency of ribosomes and
lowered activity of translational elongation factors
(54
55
56)
.
Our studies demonstrate increased methylation of the 28S region of rDNA is associated with cellular senescence. Additional experiments are necessary to clarify whether increased methylation is restricted only to the 28S region of rDNA or is a more general phenomenon. Methylation of CpG sites could potentially be either a consequence or cause of cellular senescence. Methylation of rDNA may be useful as a marker of aging both in vivo and in vitro. These connections are supported by our finding of accelerated methylation in the rDNA genes of WS fibroblasts. As WS is associated with a premature aging phenotype, an accelerated increase in the methylation of rDNA genes and other specific areas of the genome could lead to silencing of some essential genes, suboptimal expression of key proteins, and premature replicative senescence. Future studies of WS should address whether there is altered expression of certain classes of genes involved in cellular growth and senescence.
|
ACKNOWLEDGMENTS |
|---|
Received for publication October 26, 1999.
Revision received March 6, 2000.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  S. Caburet, C. Conti, C. Schurra, R. Lebofsky, S. J. Edelstein, and A. Bensimon Human ribosomal RNA gene arrays display a broad range of palindromic structures Genome Res., August 1, 2005; 15(8): 1079 - 1085. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. Ghoshal, S. Majumder, J. Datta, T. Motiwala, S. Bai, S. M. Sharma, W. Frankel, and S. T. Jacob Role of Human Ribosomal RNA (rRNA) Promoter Methylation and of Methyl-CpG-binding Protein MBD2 in the Suppression of rRNA Gene Expression J. Biol. Chem., February 20, 2004; 279(8): 6783 - 6793. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 O. Fisher, R. Siman-Tov, and S. Ankri Characterization of cytosine methylated regions and 5-cytosine DNA methyltransferase (Ehmeth) in the protozoan parasite Entamoeba histolytica Nucleic Acids Res., January 9, 2004; 32(1): 287 - 297. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. L. Opresko, W.-H. Cheng, C. von Kobbe, J. A. Harrigan, and V. A. Bohr Werner syndrome and the function of the Werner protein; what they can teach us about the molecular aging process. Carcinogenesis, May 1, 2003; 24(5): 791 - 802. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Fry The Werner Syndrome Helicase-Nuclease--One Protein, Many Mysteries Sci. Aging Knowl. Environ., April 3, 2002; 2002(13): re2 - 2. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
DNA) are indicated to the right of
the bottom right panel. C) A direct comparison of the
Southern blots of PDL#27 of AG03141 (WS), and PDL#52 of AG05247
(87-year-old donor), highlighting the slower migrating rDNA species
only observed in samples from senescent cultures of old and WS
fibroblasts.
