| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Departments of Pathology, Beth Israel Deaconess Medical Center and Harvard Medical School, Boston, Massachusetts 02215, USA
1Correspondence: Beth Israel Deaconess Medical Center, Department of Pathology, RN 280 D, 330 Brookline Ave., Boston, MA 02215, USA. E-mail: evasile{at}caregroup.harvard.edu
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Key Words: cDNA arrays • lipoproteins • human aorta
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
, 2)
. Senescence is not to be confused with
cell death in that senescent cells can survive and remain metabolically
active for long periods (3)
. Replicative senescence can be
delayed and, in some instances, reversed by genetic modifications such
as transfection with SV40 large T antigen (4
, 5)
,
activation of the telomerase gene (6
, 7)
, inactivation of
p53 (8)
, or by culture with various cytokines
(12
13
14)
. It has also been possible to identify senescent
human cells in vivo based on their morphology and expression
of neutral ß galactosidase (9
, 10)
.
In recent years, senescence has become an increasingly important topic
of scientific investigation because of its relevance to such
age-related diseases as atherosclerosis, Alzheimer’s disease, diabetes
and the more general problem of organismic aging. While generally
regarded as disadvantageous to the host, senescence may in some
instances be of benefit as, for example, in limiting the division of
premalignant cells and therefore delaying tumorigenesis. Also,
anticancer agents have been reported to induce a senescence-like
phenotype that may provide an important evaluation tool for assessing
treatment efficacy (9
10
11)
.
Like other mammalian cells, endothelial cells (EC) have a finite
proliferative life span in culture. After 
20–25 population
doublings, a substantial fraction of cells become enlarged and are
often multinucleated with a veil-like cytoplasm that is intensely
positive for senescence-associated neutral ß-galactosidase (SA
ß-Gal). Such EC may survive in a nonproliferative mode for at least
several months (10
, 12)
. Moreover, the onset of EC
senescence can be delayed or prevented by certain growth factors,
including VPF/VEGF, TGF-ß, IGF, or IL-1
(12
13
14)
.
Although it is certain that endothelial cells undergo replicative
senescence in culture, it has not been shown that they do so in
vivo. The EC lining both large and small vessels of normal adults
are quiescent cells that seldom divide and turn over very slowly
(15
, 16)
. Little is known about the capacity of large
vessel (e.g., aorta) endothelium to replicate in vivo.
However, microvascular EC are not to be regarded as senescent because
they retain the capacity to divide rapidly in response to cytokines and
growth factors emitted by tumors, wounding, inflammation, etc.
(17
18
19)
. Nonetheless, there have been suggestions of
slowed microvascular EC replication in vivo with aging.
Thus, the angiogenic response to tumors and ischemia is impaired in
older animals and can be corrected, at least in part, by administration
of VPF/VEGF (20
21
22)
. This last finding is of particular
interest because of VPF/VEGF’s ability to regulate EC gene expression
and delay EC senescence in vitro.
The goal of this study was twofold: 1) to investigate
altered gene expression in senescent vascular endothelium and
2) to determine whether vascular endothelium undergoes
senescence in vivo. We report here that early passage human
dermal endothelial cells (HDMEC) cultured with or without VPF/VEGF
overexpressed 9 genes and underexpressed 6 others in comparison with
their senescent counterparts. Of particular interest was the finding
that VPF/VEGF regulates thymosin ß-10 expression and that thymosin
ß-10 was strikingly down-regulated in senescent EC. The ß-thymosins
are a large family of monomeric actin-sequestering peptides that
regulate cell proliferation (23)
and actin dynamics,
important processes in cell migration, angiogenesis, neurogenesis, etc.
(24
, 25)
. With regard to the second goal, we found
that EC overlying atherosclerotic lesions in the aorta were strongly SA
ß-Gal-positive but reacted only weakly with antibody to thymosin
ß-10, i.e., they exhibited the senescence phenotype.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
) were from R&D Systems (Minneapolis, Minn.). All
other tissue culture reagents were purchased from Gibco-BRL.
3,3'Diaminobenzidine tetrachloride (DAB) developing color reagent was
purchased from Zymed Laboratories (San Francisco, Calif.). SuperSignal
Chemiluminescent System was from Pierce (Richmond, Ill.); biotinylated
Ulex Europaeus agglutinin-1 and Vectashield were from Vector
Laboratories (Burlingame, Calif.). Other reagents were from Sigma (St.
Louis, Mo.).
Antibodies
Rabbit anti thymosin ß-10 antibody (serum) No.1 was a generous
gift of Dr. Helen Yin (University of Texas Southwestern Medical School,
Dallas, Tex.). To generate anti-thymosin ß-10 No.2 antibody, the
carboxyl-terminal thymosin ß-10 specific peptide TIEQEKRSEIS was
synthesized, coupled to keyhole limpet hemocyanine (KLH) (Genemed
Synthesis Inc, San Francisco, Calif.) and injected into New Zealand
White rabbits (Lampire Biological Laboratories, Pipersville, Pa.). The
rabbit preimmune and immune IgG were purified by protein A Sepharose
chromatography. Other antibodies used in this study were mouse
anti-human endothelial leukocyte adhesion molecule-1 (ELAM-1/CD62E),
mouse monoclonal anti-human platelet endothelial cell adhesion molecule
(PECAM/CD31) from R&D Systems; rabbit anti-human CD31, a generous gift
from Dr. H. M. Delisser (University of Pennsylvania, Medical
Center, Philadelphia), Dynabeads M-450-coated with goat anti-mouse IgG
from Dynal (Lake Success, N.Y.). Fluorescein isothiocyanate (FITC) and
Texas red conjugated secondary antibodies were obtained from Cappel
Organon Teknika (Durham, N.C.), donkey anti-rabbit IgG coupled to Cy5
was purchased from Jackson Immunoresearch Laboratories (West Grove,
Pa.); biotinylated goat anti-rabbit IgG, biotinylated goat anti-mouse
IgG (Vectastain ABC kits) were from Vector Laboratories.
Endothelial cell Isolation and culture
Human dermal microvascular endothelial cells (HDMEC) were
isolated from human neonatal foreskins (26)
. Three to five
foreskins were pooled, finely minced, and digested with collagenase
type XI (5 mg/ml, Sigma) for 60 min at 37°C. The tissue digest was
filtered through a 70 µm nylon filter (Falcon) and the filtrate
centrifuged at 1000 g. Pelleted cells and cell clusters were
further dissociated with trypsin (0.05%, 15 min at 37°C), incubated
with 10% serum-containing media, and recentrifuged. Cells were then
plated on collagen I (50 µg/ml) -coated plates in EGM containing 10%
fetal bovine serum, 25 ng/ml cyclic adenosine monophosphate, 1 µg/ml
hydrocortisone and antibiotics. Two to three days later, EC were
selectively activated with TNF- (100 ng/ml, 4 h) and recovered
apart from contaminating cells with ELAM-1-coated dynabeads. This step
was repeated twice in order to completely eliminate nonendothelial cell
contaminants. HDMEC purity was judged by morphology, expression of CD31
antigen and by the uptake of DiI-acetylated LDL (DiI-Ac-LDL).
Endothelial cells were maintained in tissue culture with or without 50
ng/ml VEGF. The cells were serially propagated in tissue culture in
such a manner that, with each passage, cells increased by two
population doublings (PD). HDMEC cultured without VPF/VEGF became
replicative senescent after PD 24–26.
Low density lipoproteins
Low density lipoproteins (LDL) were separated from human plasma
as described previously (27
, 28)
. LDL were extensively
dialyzed against 0.15 M NaCl solution containing 1 mM EDTA, 1 mM
butylated hydroxy-toluene (BHT). Protein content was measured by a
modified Lowry assay (29)
. LDL was then fluorescently
labeled with 1,1'-dioctadecyl 3,3' tetramethyl indocarbocyanine
perchlorate (DiI-LDL) as described (30)
. DiI-oxidized LDL
(DiI-ox-LDL) was prepared by dialysis against 0.1 M NaCl solution, pH
7.4, containing 5 µM CuSO4 for 16 h at 4°C. Oxidation was
terminated by dialysis against 0.1 M NaCl, pH 7.4, supplemented with 5
mM EDTA and 0.1 mM BHT. DiI-acetylated-LDL (DiI-Ac-LDL) was prepared as
described previously (31)
or was purchased from Biomedical
Technologies Inc. (Stoughton, Mass.). LDL was radio-iodinated by the
ICl method as modified by Bilheimer et al., (32)
. Specific
activity ranged from 120 to 250 cpm/ng LDL protein.
125I-oxLDL uptake and degradation assays
HDMEC (PD12 and PD26) were plated at 250 K cells per well in
6-well dishes and maintained in EGM for 48 h, after which cells
were incubated in EBM containing 3% (v/v) newborn calf
lipoprotein-deficient serum for 16 h. Thereafter, the cells were
rinsed with fresh serum-free medium and cultured for 5 h at 37°C
with fresh medium supplemented with 10 µg/ml
125I-oxLDL in the absence (duplicate
determinations) or presence of 400 µg/ml unlabeled ox-LDL (single
determinations). Culture supernatants were recovered and
125I-oxLDL degradation products released by the
cells into the media were measured as described previously (31
, 33)
. In brief, culture supernatants were precipitated with 20%
trichloro-acetic acid (TCA); TCA supernatants were treated with
potassium iodide and H2O2
and then extracted with chloroform to remove free iodine. Aliquots of
the aqueous phase of the chloroform extracted supernatants were
subjected to gamma counting to provide a measure of degraded ox-LDL
expressed as nanograms of 125I-ox-LDL degraded
per milligram of cell protein in 5 h. Cell monolayers were rinsed
with cold phosphate-buffered saline (PBS) and dissolved in 0.1 N NaOH.
Aliquots of cell extracts were used to measure cell protein and the
remainder was subjected to gamma counting to measure cell uptake of
125I-oxLDL.
Senescence associated ß-galactosidase (SA ß-Gal) staining
Cultured cells were rinsed in PBS, fixed in 4% formaldehyde in
PBS for 4 min at room temperature, then briefly washed in PBS and
incubated with ß Gal solution (pH 6.0) for 8–16 h at 37°C
(10)
.
Human aorta and skin samples
Portions of human aortic arch exhibiting atherosclerotic lesions
were collected from three patients, all older than 70 years, at autopsy
(18–24 h postmortem) and fixed in 4% formaldehyde in PBS for 1 h. Segments of aorta were subjected to SA ß-Gal staining directly as
above. Portions of blue staining and adjacent nonstaining (normal)
aortic arch were snap-frozen in Tissue Tek O.C.T. embedding medium (Ted
Pela, Reading, Calif.).
Immunohistochemistry
Frozen sections (10 µm thick) of blue staining and
normal-appearing aortic arch were incubated in PBS containing 0.3%
hydrogen peroxide for 30 min to quench tissue endogenous peroxidases,
rinsed in PBS, and then incubated with 50 mM sodium borohydride for 30
min to quench the nonspecific binding and autofluorescence of elastin.
To block nonspecific binding and to permeabilize the tissue, the
sections were further incubated in PBS containing 1% bovine serum
albumin (BSA) and 0.1% saponin at 4°C, overnight. Sequential tissue
sections were then incubated with mouse monoclonal anti-human CD31 (10
µg/ml) or rabbit anti-human thymosin IgG (1 µg/ml) (No.2 antibody).
As negative controls, normal mouse IgG, normal rabbit IgG, preimmune
rabbit IgG (serum collected before immunizing the rabbits with thymosin
ß-10) were used instead of the first antibody at the same protein
concentration as the corresponding primary antibodies or the first
antibody was omitted. Secondary antibodies (goat anti-mouse or goat
anti-rabbit biotinylated IgG, respectively) were then applied followed
by avidin-peroxidase according to the manufacturer’s protocol
(Vector). The peroxidase reaction was developed using DAB plus kit for
5 min. CD-31 antigen detection on aortic endothelial layer was further
amplified by using DAB enhancer according to the manufacturer protocol
(Zymed Laboratories). Other sections were reacted with biotinylated
Ulex Europaeus agglutinin-1 and developed with avidin-peroxidase and
AEC substrate (Zymed Laboratories). The slides were washed, dehydrated
and mounted with Cytoseal mounting medium (Fisher Scientific,
Pittsburgh, Pa.), viewed and photographed with a Zeiss Axiophot
microscope system and Kodak Ektachrome film.
Immunofluorescence
PD 12 and PD 26 HDMEC grown on collagen-coated coverslips were
fixed with 4% formaldehyde in PBS for 1 h at room temperature.
Cells were then incubated for 30 min with PBS containing 1% BSA to
block nonspecific binding and were permeabilized in PBS - 0.1% saponin
- 1% BSA for 30 min. Cells were then incubated with rabbit
anti-thymosin ß-10 antibody (1: 200 serum No.1 or 1 µg/ml IgG
obtained from serum No.2), washed in PBS, and incubated with 1: 500
goat anti-rabbit FITC-IgG (secondary antibody). After further washing
with PBS, the cells were mounted on microscope slides with Vectashield
(Vector). A similar procedure was used for aortic frozen fixed sections
(see Immunohistochemistry above). As negative controls, normal rabbit
IgG and preimmune serum were used instead of the first antibody or the
first antibody was omitted (34)
.). For aortic sections, to
avoid any elastin blue autofluorescence the secondary antibody was
donkey anti-rabbit IgG coupled to Cy5 (far-red fluorescence
fluorophore). Cell nuclei were stained with propidium iodide. The
samples were viewed with a Bio-Rad MRC-1024 confocal microscope. For
photographic reproduction, the Cy5 far-red fluorescence was converted
to a green display with Bio-Rad Lasersharp RGB merge software.
Differential gene expression
Poly A+ RNAs were isolated from PD 12 and
PD 26 HDMEC cultured with or without VPF/VEGF from the time of
isolation using a commercial kit (Clontech, Palo Alto, Calif.).
32P-labeled ATP-cDNA probes were generated by
reverse transcription of purified poly(A)+ RNAs.
cDNA probes prepared from PD12 and PD 26 HDMEC were hybridized in
parallel on identical cDNA array membranes (Human Atlas Array I,
Clontech Laboratories Inc., Palo Alto, Calif.) according to the
manufacturer’s instructions (Clontech # PT3140–1). Each membrane
contained 588 human cDNAs spotted in duplicate. Plasmid and
bacteriophage DNAs were included as negative controls to confirm
hybridization specificity along with nine housekeeping cDNAs as
positive controls for normalizing mRNA abundance. The spotted cDNAs
were arrayed into six functional classes: oncogenes, stress response,
apoptosis, transcription factors, receptors, and cytoskeleton motility
genes. The hybridized membranes were exposed to BioMax MS film (with a
BioMax TranScreen intensifying screen), (Eastman Kodak, New Haven,
Conn.) at -70°C for 5–7 days.
Northern blotting
Total RNA and mRNA were extracted from PD 12 and PD 26 HDMEC
cultured with or without VPF/VEGF using a Qiagen kit according to
manufacturer’s protocol (Qiagen, Valencia, Calif.). Northern blots
were hybridized with a 32P-labeled CTP
thymosin-ß 10 probe that was generated by RT-PCR from a human uterus
cDNA library using oligonucleotide primers (Clontech, Palo Alto,
Calif.). A probe specific for the ribosome associated protein 36B4 was
used to control for RNA loading. Hybridized blots were exposed to
BioMax MS film that was scanned at 100 µm pixel size using a
Molecular Dynamics densitometer (Sunnyvale, Calif.). Digitized images
were analyzed by ImageQuant software (Molecular Dynamics) using ‘the
local background’ correction method.
In situ hybridization
Human thymosin-ß 10 cDNA (300 bp, spanning from 40 to 340 bp
of 400 bp entire coding region) was subcloned into a pGEM-1 vector
(Promega, Madison, Wis.) from which single-stranded (sense and
antisense) RNA probes were synthesized (108
cpm/µg) with 35S-UTP, purified on
polyacrylamide gels, and used without reduction in length. In
situ hybridization was performed as described previously
(35)
.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
, VPF/VEGF delayed the
onset of senescence in HDMEC. HDMEC cultured without VPF/VEGF exhibited
growth arrest or replicative senescence by PD 24–26 whereas HDMEC
cultured continuously with VPF/VEGF continued to proliferate beyond PD
38–40. To investigate the underlying mechanisms, we made use of cDNA
expression arrays to compare the genes expressed by early passage
PD12-HDMEC cultured with VPF/VEGF with those expressed by
growth-arrested, senescent PD 26 HDMEC cultured without VPF/VEGF.
Autoradiographic analysis of paired cDNA expression arrays revealed
that 9 genes were relatively overexpressed by PD 12 HDMEC cultured with
VPF/VEGF whereas 6 different genes were relatively overexpressed by
senescent PD 26 HDMEC (Fig. 1
). Genes relatively overexpressed by PD 12 HDMEC included Tie-1,
thrombin receptor, integrin -5, -catenin, endothelin, HSP 27,
monocyte chemotactic and activating factor (MCAF), ICAM-1, and thymosin
ß-10 genes. Significantly overexpressed by senescent PD 26 HDMEC were
growth arrest and DNA damage-inducible protein (GADD) 153, integrin ß
1, PDGF receptor-, basic fibroblast growth factor (bFGF) receptor,
DNA binding protein TAXREB67, and nuclease sensitive DNA binding
protein genes. Both of the high-affinity VPF/VEGF receptors, VEGFR-1
(flt-1) and VEGFR-2 (KDR), were expressed equally and at low levels by
both PD 12 and PD 26 HDMEC. Thymosin ß-10 was the most up-regulated
gene in early passage as compared to senescent/late passage HDMEC. Gene
expression in PD12 HDMEC cultured with or without VPF/VEGF was similar
except that Tie-1 was expressed relatively less in cells cultured
without VPF/VEGF. HDMEC cultured continuously with VPF/VEGF through PD
26 did not exhibit the senescence phenotype.
|
Relative overexpression of thymosin ß-10 transcript by PD 12 HDMEC
was confirmed by Northern blotting (Fig. 2A
) and was found to be three- to fourfold more abundant in PD
12 HDMEC (cultured with or without VPF/VEGF) than in senescent PD 26
cells (Fig. 2B
). As expected from analysis of cDNA
expression arrays, addition of exogenous VPF/VEGF to culture medium
induced only a small additional increase in the expression of thymosin
ß-10 in early passage (PD 12) HDMEC (Fig. 2)
; however, HDMEC cultured
continuously with VPF/VEGF from initial seeding through PD 26
expressed twofold more thymosin ß-10 than PD 26 HDMEC
cultured without VPF/VEGF (Fig. 2B
, compare lane 3 and 4).
|
Cellular distribution of thymosin ß-10 protein
The distribution of thymosin ß-10 was compared in early passage
PD12 HDMEC and in senescent PD26 HDMEC by confocal fluorescence
microscopy. PD 12 HDMEC, cultured with or without VPF/VEGF, exhibited
intense, diffuse cytoplasmic staining for thymosin ß-10 (Fig. 3A
, B
). Staining was considerably weaker in PD 26 HDMEC
cultured without VPF/VEGF and exhibiting the senescence phenotype
(enlarged, multinucleated, with veil-like cytoplasm) (Fig. 3C
). PD 26 HDMEC grown with VPF/VEGF did not become
senescent and their thymosin ß-10 staining was similar to that of PD
12 HDMEC (data not shown).
|
Immunohistochemical detection of senescent cells in human aorta
To determine whether senescent endothelial cells were present
in situ, segments of human aorta obtained at autopsy from
elderly patients were processed for SA ß-Gal staining. Zones (1–4 mm
in diameter) of characteristic blue luminal staining developed over
atherosclerotic plaques on the aortic arch and adjacent to the carotid
bifurcation (Fig. 4A
, B
). This finding was confirmed by histochemistry as a
thin, continuous layer of luminal SA ß-Gal staining; scattered cells
located immediately beneath the endothelial layer (possibly
representing macrophages or smooth muscle cell) also stained (Fig. 4C
).
|
Next we compared the distribution of thymosin ß-10 in
SA-ß-Gal-positive and -negative regions of aortic arch (Fig. 5
). The SA ß-Gal-positive endothelial layer overlying atherosclerotic
lesion stained less intensely for thymosin ß-10 than adjacent
ß-Gal-negative endothelium (Fig. 5A
, B
). Immunostaining
with CD31 confirmed that the SA-ß-Gal-positive area of staining of
the aortic lumen in fact represented endothelium (Fig. 5C
);
a similar thin layer of CD31 immunostaining was also observed in
SA-ß-Gal-negative portions of the aorta. Endothelial cells lining
small adventitial vessels in the aortic adventitia stained strongly for
thymosin ß-10 (Fig. 5D
). Taken together, these results
demonstrate that EC overlying atherosclerotic plaques exhibit
characteristics of senescence (strong staining for SA-ß-Gal and weak
staining for thymosin ß-10).
|
Uptake of modified lipoproteins by HDMEC
Oxidized LDL are thought to play an important role in
atherosclerosis and may contribute to arterial wall inflammation and
cellular injury (36)
. Both ox-LDL and ac-LDL bind to a set
of functionally overlapping receptors expressed on vascular
endothelium. They are subsequently internalized by receptor-mediated
endocytosis and are then degraded in lysosomes (36
, 37)
.
At PD 20, up to 10% of the HDMEC population cultured without VPF/VEGF
exhibited the characteristic cell enlargement and SA-ß-Gal positivity
of senescence. When PD 20 cells were incubated with DiI-ox-LDL or
DiI-ac-LDL (10 µg/ml) for 5 h at 37°C, cells exhibiting the
senescence phenotype exhibited more extensive uptake of both tracers
than smaller, normal-sized nonsenescent endothelial cells (Fig. 6A
, B
).
|
To quantitate lipoprotein uptake and degradation by early passage and senescent HDMEC, we incubated PD12 and senescent PD 26 HDMEC (both cultured without VPF/VEGF) with 125I-ox-LDL. After 5 h at 37°C, similar amounts of 125I ox-LDL per milligram cell protein were found in PD 12 and PD 26 HDMEC. In a typical experiment (one of three), early passage PD 12 HDMEC contained 563 ± 39.5 ng 125I ox-LDL/mg cell protein vs. 548 ± 42.5 ng 125I ox-LDL/mg cell protein in senescent PD26 HDMEC. However, lipoprotein degradation was more than threefold greater in PD12 HDMEC than in PD 26 HDMEC: 446 ± 2.2 ng protein I125 ox-LDL/mg cell protein/5 h in culture medium of PD 12 vs. 130 ± 11.5 ng protein 125I ox-LDL/mg cell protein/5 h in culture medium of PD26 HDMEC. These results demonstrate that although individual enlarged senescent HDMEC stored larger amounts of lipoproteins than individual early passage cells, the content of lipoprotein did not differ in senescent and early passage HDMEC when expressed as lipoprotein per milligram cell protein. Furthermore, the data indicate that atherogenic oxidized lipoprotein degradation is greatly reduced in senescent endothelium.
Detection of thymosin ß-10 in human adenocarcinomas of the colon
and in the tumor microvasculature
Thymosin ß-10 is expressed in many tissues and in increased
amounts in certain tumors including melanomas and carcinomas of the
kidney, ovary, and medullary carcinoma of the breast (23
, 38)
. We have now extended this finding to another important
human tumor, demonstrating that adenocarcinomas of the colon express
much more thymosin ß-10 than adjacent normal epithelium (Fig. 7A
, B
). Given the strong expression of thymosin ß-10 in
actively dividing, early passage microvascular endothelium, we
hypothesized that thymosin ß-10 might also be strongly expressed in
tumor blood vessels. In fact, high levels of thymosin ß-10 mRNA were
found by in situ hybridization in endothelial cells lining
tumor microvessels (Fig. 7E
); however, thymosin ß-10 mRNA
was also expressed strongly by vascular endothelium in adjacent,
nonneoplastic tissue sections (Fig. 7E
). Similar results
were obtained by immunohistochemical staining with an antibody to
thymosin ß-10.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
. The fact that
early passage HDMEC expressed large and equivalent amounts of thymosin
ß-10 in the presence or absence of exogenous VPF/VEGF suggests that
VPF/VEGF is not necessary for maintaining thymosin ß-10 expression in
the short term. Very likely, early passages EC already express maximal
levels of thymosin ß-10. However, VPF/VEGF did increase thymosin
ß-10 expression in late passage presenescent HDMEC, which expressed
much lower levels of thymosin ß-10 than early passage cells.
Thymosin ß-10 is an actin G-sequestering protein that is highly
expressed in embryonic and tumor cells (23
, 38
, 49
, 50)
, a
finding we have extended here to adenocarcinomas of the colon. It is
thought to regulate the dynamics of the actin cytoskeleton and thereby
modulate cell motility (51
, 52)
and its overexpression is
reported to accelerate apoptosis (53)
. The great majority
of malignant tumors overexpress both VPF/VEGF and thymosin ß-10 and
have a proclivity for apoptosis. The mechanisms by which added VPF/VEGF
maintains thymosin ß-10 expression and delays senescence in cultured
HDMEC are not well understood and require further investigation.
A second novel finding was that the aortic endothelium overlying
atherosclerotic plaques exhibited properties of senescence, i.e., were
SA ß-Gal-positive and stained only weakly for thymosin ß-10. In
contrast, the endothelium covering nearby regions of relatively normal
aorta were SA ß-Gal-negative and stained strongly for thymosin
ß-10. These observations are consistent with earlier work performed
on the aortic endothelium of hyperlipidemic rabbits that demonstrated
that endothelial cells overlying developing atherosclerotic plaques
were greatly enlarged and lipid filled, thus presenting a phenotype
consistent with the senescence reported here (41
, 42)
.
Similarly, enlarged multinucleated EC have been reported overlying
advanced plaques in human atherosclerosis (43)
. Other
investigators have reported that cultured endothelial cells treated
with oxidized LDL exhibited some morphological characteristics
(blebbing and marked loss of viability) similar to those of EC covering
atherosclerotic areas (44)
. Together, these data provide
strong evidence that endothelial cells overlying atherosclerotic
plaques undergo senescence and therefore the large vessel vascular
endothelium has the capacity to undergo senescence in vivo
as well as in vitro.
A third finding concerned differences in turnover of DiI-ac-LDL and
DiI-ox-LDL by senescent and nonsenescent endothelium. Both early
passage and senescent HDMEC endocytosed large amounts of DiI-ac LDL and
DiI-ox-LDL after 5 h of incubation. Individual large, senescent
HDMEC contained substantially more fluorescent lipoprotein than their
smaller, nonsenescent counterparts. This finding likely reflects the
larger lysosomal compartment that has been described in senescent
endothelial cells (39
, 40)
. However, when considered in
terms of lipoprotein incorporated per milligram EC protein, no
difference was found in the amounts of lipoprotein incorporated by
senescent or nonsenescent HDMEC. We did, however, find that early
passage EC degraded almost threefold more lipoprotein than senescent
cells over the same period of time, indicating a substantial reduction
in the capacity of senescent EC to metabolize atherogenic lipids.
Oxidized LDL are known to stimulate EC proliferation, and perhaps
thereby the aging process (45
, 46)
and directly
up-regulate the expression of stress-related proteins, such as heat
shock protein 70 (HSP 70) in human umbilical vein endothelial cells
(HUVEC) (47)
. We found that one stress-related protein,
HSP 27, was slightly up-regulated in actively dividing early passage
HMDEC grown in the presence of the VPF/VEGF as compared with senescent
cells.
Senescence-associated gene expression was also measured recently in
HUVEC by microarray analysis (48)
. Of the 12 genes that
were up-regulated in early passage HUVEC, only four (cyclins A and B1,
cdc2, and HSP 70B) were represented in our array, and we found that
none of these was up-regulated in early passage vs. senescent HDMEC.
However, we did find that another heat shock protein, HSP 27, was
up-regulated in PD 12 HDMEC. Of the 17 genes that Shelton et al.
(48)
found up-regulated in senescent HUVEC, only 4 were
represented in our array: one of these, GADD 153, was also relatively
overexpressed in senescent HDMEC; levels of two (adenosine A2A receptor
and TGF-ß 2) were unchanged; and one (ICAM-1) was actually
down-regulated in senescent as compared with early passage HUVEC. These
data indicate that even in cells of the same general type (vascular
endothelium), large differences in gene expression are to be expected
when cells are isolated from different vascular beds (in this case,
HDMEC vs. HUVEC) and that environmental factors (e.g., culture
conditions) may significantly affect gene expression in addition to
passage doubling number. These findings emphasize the importance of
demonstrating differences in gene expression in vivo (here
for thymosin ß-10) as well as in vitro.
|
ACKNOWLEDGMENTS |
|---|
Received for publication February 1, 2000.
Revision received May 15, 2000.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  A. S. Multani and S. Chang WRN at telomeres: implications for aging and cancer J. Cell Sci., March 1, 2007; 120(5): 713 - 721. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. Minamino and I. Komuro Vascular Cell Senescence: Contribution to Atherosclerosis Circ. Res., January 5, 2007; 100(1): 15 - 26. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J.-P. Coppe, K. Kauser, J. Campisi, and C. M. Beausejour Secretion of Vascular Endothelial Growth Factor by Primary Human Fibroblasts at Senescence J. Biol. Chem., October 6, 2006; 281(40): 29568 - 29574. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 I. Gorenne, M. Kavurma, S. Scott, and M. Bennett Vascular smooth muscle cell senescence in atherosclerosis Cardiovasc Res, October 1, 2006; 72(1): 9 - 17. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. Kunieda, T. Minamino, J.-i. Nishi, K. Tateno, T. Oyama, T. Katsuno, H. Miyauchi, M. Orimo, S. Okada, M. Takamura, et al. Angiotensin II Induces Premature Senescence of Vascular Smooth Muscle Cells and Accelerates the Development of Atherosclerosis via a p21-Dependent Pathway Circulation, August 29, 2006; 114(9): 953 - 960. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. Matthews, I. Gorenne, S. Scott, N. Figg, P. Kirkpatrick, A. Ritchie, M. Goddard, and M. Bennett Vascular Smooth Muscle Cells Undergo Telomere-Based Senescence in Human Atherosclerosis: Effects of Telomerase and Oxidative Stress Circ. Res., July 21, 2006; 99(2): 156 - 164. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Parrinello, J.-P. Coppe, A. Krtolica, and J. Campisi Stromal-epithelial interactions in aging and cancer: senescent fibroblasts alter epithelial cell differentiation J. Cell Sci., February 1, 2005; 118(3): 485 - 496. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. O. d. Buijs, M. Musters, T. Verrips, J. A. Post, B. Braam, and N. van Riel Mathematical modeling of vascular endothelial layer maintenance: the role of endothelial cell division, progenitor cell homing, and telomere shortening Am J Physiol Heart Circ Physiol, December 1, 2004; 287(6): H2651 - H2658. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Seluanov, D. Mittelman, O. M. Pereira-Smith, J. H. Wilson, and V. Gorbunova DNA end joining becomes less efficient and more error-prone during cellular senescence PNAS, May 18, 2004; 101(20): 7624 - 7629. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. J. Kurz, S. Decary, Y. Hong, E. Trivier, A. Akhmedov, and J. D. Erusalimsky Chronic oxidative stress compromises telomere integrity and accelerates the onset of senescence in human endothelial cells J. Cell Sci., May 1, 2004; 117(11): 2417 - 2426. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 B. T. Weinert and P. S. Timiras Invited Review: Theories of aging J Appl Physiol, October 1, 2003; 95(4): 1706 - 1716. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. R. Smith Author's Response to Commentary: Cell Senescence: An Evaluation of Replicative Senescence in Culture as a Model for Aging In Situ J. Gerontol. A Biol. Sci. Med. Sci., September 1, 2003; 58(9): B779 - 781. [Full Text] [PDF]
|
|
|
|
 |
