Content uploaded by Oksana Kehoe
Author content
All content in this area was uploaded by Oksana Kehoe on Jul 16, 2014
Content may be subject to copyright.
The Angiopoietin/Tie-2 System Regulates Pericyte
Survival and Recruitment in Diabetic Retinopathy
Jun Cai,
1
Oksana Kehoe,
2
Gill M. Smith,
2
Philip Hykin,
3
and Michael E. Boulton
1
PURPOSE. The angiopoietin (Ang) system plays an important
role in vascular stabilization and pathologic neovascularization.
The hypothesis for the study was that, in addition to modulat-
ing endothelial cell behavior, the angiopoietin/Tie-2 system
also regulates the pericyte apoptosis and/or the vessel matura-
tion associated with diabetic retinopathy.
M
ETHODS. Tie-2 expression in cultured retinal pericytes was
analyzed by using ELISA, Western Blot analysis, and flow cy-
tometry. CD13 (aminopeptidase N) expression in pericytes
was determined by Western blot analysis and Ang effects ver-
ified with Tie-2 antisense treatment. Cell proliferation was
monitored by crystal violet uptake, and pericyte migration was
assessed in a scrape wound. Annexin V-FITC flow cytometry
was used to quantify pericyte apoptosis.
R
ESULTS. Pericytes expressed a functionally active Tie-2 recep-
tor upregulated by both Ang-1 and -2 (P ⬍ 0.05). In pericytes
undergoing apoptosis induced by either TNF-
␣
or high glucose
Ang-1 increased survival (P ⬍ 0.05 for TNF-
␣
; P ⬍ 0.01 for high
glucose), whereas Ang-2 increased apoptosis. Ang-1 enhanced
CD13 expression in a dose-dependent manner (P ⬍ 0.05) and
stimulated pericyte migration in a synthetic matrix wound-
healing assay that was associated with a change in F-actin
organization. Addition of Tie-2 antisense confirmed that angio-
poietins act through Tie-2.
C
ONCLUSIONS. These findings demonstrate that Tie-2 is func-
tional in pericytes and may play an important role in the
progression of diabetic retinopathy, by regulating pericyte loss
and influencing the activation state and recruitment of
pericytes. (Invest Ophthalmol Vis Sci. 2008;49:2163–2171)
DOI:10.1167/iovs.07-1206
A
hallmark of diabetic retinopathy and a precursor of endo-
thelial cell dysfunction is the loss of pericytes, so-called
“pericyte drop-out,” leading to acellular capillaries, which are
no longer perfused.
1–3
Although the loss of pericytes is be-
lieved to be linked to fluctuations in blood glucose, the mech-
anism is not well understood. In addition, the maturation of
newly formed vessels during the late stages of proliferative
diabetic retinopathy requires the incorporation of pericytes.
4
The angiopoietins, a recently discovered family of vascular
regulatory molecules binding to the Tie-2 tyrosine receptor,
play an important role in retinal neovascularization.
5,6
The
TIE-2 gene encodes a protein of 1122 amino acids with its
extracellular region containing three EGF-like repeats and
three repeats with fibronectin type III homology located after
the second Ig loop. The intracellular portion of Tie-2 contains
two tyrosine kinase domains. Tie-2-deficient mice die at around
embryonic day 10.5 and exhibit profound vascular defects
indicating the importance of Tie-2 in vascular development.
7,8
The activity of Tie-2 is differentially regulated by angiopoietin
(Ang)-1 and -2, which share approximately 60% amino acid
identity with similar affinity to Tie-2.
9,10
Although primarily
thought of as an endothelial cell protein, Tie-2 is also expressed
by some hematopoietic stem cells,
11
and its RNA has been
reported in pericytes.
12
Ang-1 binds to Tie-2 and induces phosphorylation of Tie-2.
Ang-1/Tie-2 are proposed to mediate the mobilization of hema-
topoietic stem cells to the peripheral circulation
13
and the
formation of mature capillary networks by recruiting perien-
dothelial cells such as pericytes.
14
Both Ang-1 and Tie-2 are
upregulated during the early stages of wound healing coincid-
ing with wound angiogenesis
15
and in various tumors and
tumor cell lines.
16
Furthermore, the constitutive secretion of
Ang-1 by normal quiescent vessels is considered to stabilize
vessels by maintaining contacts between endothelial cells and
periendothelial cells.
Ang-2 has been suggested to block the constitutive stabili-
zation or maturation function of Ang-1 by promoting smooth
muscle cell–pericyte dropout therefore loosening contacts be-
tween endothelial cells and periendothelial cells.
17
Ang-2 has
been reported to be upregulated in mouse models of ischemia-
induced retinal neovascularization, as well as during angiogen-
esis in retinal development.
18
We investigated how the angiopoietin signaling system may
regulate the course of diabetic retinopathy by modulating ret-
inal pericyte survival and recruitment.
METHODS
Cell Culture
Retinal pericytes were isolated from bovine eyes and cultured by using
a modification of previous methods.
19
Isolated pericytes were resus-
pended in MEM with 20% fetal bovine serum at 37°C for 3 days. After
cell attachment, the medium was changed every 3 to 4 days with MEM
containing 10% fetal bovine serum.
Purification of bovine retinal pericytes was achieved with a kit
(Cellection Pan Mouse IgG kit; Dynal Biotech, Bromborough, UK)
according to the manufacturer’s instructions. Briefly, the cells col-
lected from one primary culture T-25-cm
2
flask by trypsinization were
reacted with mouse anti-desmin monoclonal antibody (Chemicon Eu-
rope, Chandlers Ford, UK) for 10 minutes at 4°C. After washing, the
cells were mixed with pan mouse IgG beads (Cellection Dynabeads;
Dynal Biotech) for 20 minutes at 4°C with constant shaking. The
From the
1
Department of Ophthalmology and Visual Sciences,
The University of Texas Medical Branch, Galveston, Texas; the
2
School
of Optometry and Vision Sciences, Cardiff University, Cardiff, Wales,
United Kingdom; and
3
Moorfields Eye Hospital, London, United King-
dom.
Supported by Diabetes UK, the Wellcome Trust, UK, and Research
to Prevent Blindness; and Regeneron Pharmaceuticals, Inc. (Tarry-
town, NY), which provided Ang-1 and -2.
Submitted for publication September 17, 2007; revised November
14, 2007, and January 4, 2008; accepted March 6, 2008.
Disclosure: J. Cai, Regeneron Pharmaceuticals, Inc. (F); O. Ke-
hoe, None; G.M. Smith, None; P. Hykin, None; M.E. Boulton,
Regeneron Pharmaceuticals, Inc. (F)
The publication costs of this article were defrayed in part by page
charge payment. This article must therefore be marked “advertise-
ment” in accordance with 18 U.S.C. §1734 solely to indicate this fact.
Corresponding author: Michael E. Boulton, Department of Oph-
thalmology and Visual Sciences, The University of Texas Medical
Branch, 301 University Boulevard, Galveston, TX 77550;
boultonm@utmb.edu.
Investigative Ophthalmology & Visual Science, May 2008, Vol. 49, No. 5
Copyright © Association for Research in Vision and Ophthalmology
2163
samples were placed in a magnetic particle concentrator (MPC-S;
Dynal Biotech) to separate immunoadsorbed cells for 2 minutes at
room temperature. The bead-bound cells were then resuspended in
release buffer for 15 minutes. After vigorous pipetting, the samples
were place in the particle concentrator for 2 minutes, and the purified
pericytes were collected and seeded into T25-cm
2
flasks under the
culture conditions just described. Cultures were subcultured at a ratio
of 1:2 on reaching confluence, and the cells were used within three
passages. For all experimental studies the fetal bovine serum concen-
tration of the medium was reduced to 5%.
To rule out endothelial cell contamination, equal amounts of lysates
from pericyte cultures and microvascular endothelial cells (as a posi-
tive control) were subjected to Western blot analysis with anti-
PECAM-1 antibody (Santa Cruz Biotechnology Inc., Santa Cruz, CA).
To confirm that pericytes are able to retain their phenotype during
progressive passage, the cells were propagated for 5 days in the second
passage and then incubated with mouse anti-desmin monoclonal anti-
body (Chemicon Europe) at room temperature for 30 minutes (mouse
IgG
1
antibody acted as a control). After incubation with an FITC-
conjugated secondary antibody the number of desmin-positive cells
was determined by flow cytometry.
To study the pericyte response to exogenous Ang-1 or -2 cells were,
unless otherwise stated, treated with Ang-1 or -2 (Regeneron Pharma-
ceuticals, Inc., Tarrytown, NY) at a concentration of 100 ng/mL, based
on preliminary experiments and the functional response achieved in
Figure 4. For the unstimulated control, Ang was substituted with 0.1%
BSA.
ELISA Determination of Tie-2
After treatment with 100 ng/mL Ang-1 or -2 at 37°C for 48 hours, the
medium was collected, and the pericytes were lysed by adding 1 mL
per flask of RIPA buffer (50 mM Tris-HCl [pH 7.4], 150 mM NaCl, 1%
NP-40, 0.25% sodium deoxycholate, 1 mM NaF, 1 mM Na
3
VO
4
and 1
mM EDTA containing the protease inhibitors phenylmethylsulfonyl
fluoride, aprotinin, leupeptin, and pepstatin) at 4°C for 30 minutes.
The amount of protein was determined by the BCA protein assay
before detection of Tie-2 by ELISA (Quantikine; R&D Systems, Europe,
Abingdon, UK) according to the manufacturer’s instructions and ab-
sorbance was read at 450 nm. The detection limit of this assay was 0.02
ng/mL. Tie-2 levels were normalized for total protein concentrations in
the cell lysates and expressed as ng Tie-2 per milligram of total protein.
Tie-2 Phosphorylation
To assess the phosphorylation status of Tie-2 after angiopoietin treat-
ment for 48 hours, phosphorylated Tie-2 was immunoprecipitated
from the cell lysate containing protein at 500
g/mL by incubation
with 10
L mouse anti-tyrosine phosphorylation monoclonal antibody
(PY 20; Santa Cruz Biotechnology, Inc.) for 1.5 hours at 4°C, followed
by addition of 20
L protein A/G agarose (Santa Cruz Biotechnology,
Inc.) overnight at 4°C. After a wash in RIPA buffer, the mixture was
centrifuged at 12,000g for 20 minutes. The pellet was resuspended
with 1 mL of RIPA buffer containing protease inhibitors at 65°C for 5
minutes, to disrupt protein–protein interactions, and was then sub-
jected to ELISA for Tie-2.
Flow Cytometric Analysis of Tie-2
Subconfluent pericytes treated with 100 ng/mL Ang-1 or -2 for 48 hours
were harvested by centrifugation at 200g for 5 minutes. After three
washes in PBS containing 0.5% FCS, the cells were resuspended in the
same buffer to a final concentration of 4 ⫻ 10
7
/mL and 25
Lofthe
cells (1 ⫻ 10
6
) were added to a tube. A 10-
L aliquot of the PE-
conjugated anti-Tie-2 antibody (R&D Systems Europe) was added to the
cells at 4°C for 45 minutes, followed by two washes in the same buffer
before flow cytometric analysis (FACSCalibur; BD Biosciences, Oxford,
UK). As a control, the cells in a separate tube were treated with
PE-labeled mouse IgG
1
antibody.
Antisense Treatment
Antisense 5⬘-GCTAAAGAATCCATGCTTCCCC-3⬘ and scrambled oligos
5⬘-GGGGGAAGCATGGATTCTTTAGC-3⬘ that corresponded to base
pairs 318 to 337 of bovine Tie-2 mRNA were constructed by MWG
Biotech (Milton Keynes, UK). In oligo transfections, 6
M of oligo,
predetermined as the optimal concentration, in MEM was incubated
with DMRIE-C for 45 minutes at 25°C. After the pericytes were
washed, liposome-oligo complexes were incubated with the cells for 5
hours at 37°C followed by addition of MEM containing 10% FCS for 19
hours. Based on the titration result, pericytes (Fig. 2C) were treated
with 6
M of the Tie-2 antisense and scrambled oligo daily over a 3-day
period.
Western Blot Analysis for CD13
CD13 (aminopeptidase N) is a marker of hematopoietic cells
20
and cell
mobility and is a putative indicator of pericyte maturation in the
cerebral vascular bed.
21
It is a type II membrane-bound metallopro
-
teinase expressed on various cell types and a Zn
2⫹
-dependent ectopep
-
tidase that degrades preferentially proteins with an NH
2
-terminal neu
-
tral amino acids.
22,23
CD13 expression in pericytes was analyzed by
Western blot analysis. Subconfluent pericytes were cultured in the
presence or the absence of 100 ng/mL Ang-1 or -2 at 37°C for 48 hours.
The proteins were extracted with RIPA buffer at 4°C, and the protein
content was determined by the BCA protein assay. Equal amounts of
protein from each sample were resolved by SDS polyacrylamide gel
and transferred onto nitrocellulose membrane. The membranes were
incubated with an antibody for CD13 (Santa Cruz Biotechnology, Inc.)
at room temperature for 2 hours. The membranes were then washed
with 3% milk/TBS containing 0.05% Tween-20 followed by HRP-con-
jugated secondary antibody (Santa Cruz Biotechnology, Inc.) at room
temperature for 1 hour. After they were washed, the membranes were
incubated with enhanced chemiluminescent substrate (ECL; Santa
Cruz Biotechnology Inc.) and exposed to film (Biomax MR; Sigma-
Aldrich, Poole, Dorset, UK). The blots were stripped and reprobed
with goat polyclonal anti-
␣
tubulin antibody (1:250; Santa Cruz Bio-
technology, Inc.).
Detection of Apoptosis
Apoptosis was evaluated by using a FITC-conjugated annexin V/pro-
pidium iodide (PI) assay kit (R&D Systems Europe), based on annexin-V
binding to phosphatidylserine exposed on the outer leaflet of the
plasma membrane lipid bilayer of cells entering the apoptotic pathway.
Apoptosis was induced using two approaches, TNF-
␣
and high glu-
cose. Pericytes were exposed to (1) 100 ng/mL Ang-1 or -2 in the
presence or absence of 100 ng/mL TNF-
␣
at 37°C for 48 hours or (2)
high glucose, 15 and 25 mM, or control medium containing 5.6 mM
glucose. Glucose levels were replenished daily. Ang-1 or -2 (100 ng/
mL) was added to the culture medium 48 hours before cell harvest.
For apoptosis detection, the cells were collected by centrifugation
(200g for 5 minutes), washed in PBS, and resuspended in the annexin
V incubation reagent in the dark for 15 minutes before flow cytometric
analysis.
Proliferation Assay
Cells were plated in 96-well plates at 750 cells per well in MEM with
10% FCS for 24 hours. They were transferred to serum-free MEM for 45
minutes and then treated with various concentrations of Ang-1 or -2 in
MEM containing 1% FCS at 37°C for 48 hours. The relative cell number
was determined by crystal violet uptake.
24
To assess the degree of cell
cycle synchronicity within the population of pericytes, the cultures
were stimulated with Ang-1 or -2 (100 ng/mL) for 48 hours and then
fixed with 70% ethanol. After digestion with RNase, DNA was stained
with propidium iodide and DNA density determined by flow cytom-
etry, to obtain a measure of cell cycle progression.
2164 Cai et al. IOVS, May 2008, Vol. 49, No. 5
Migration Assay
In vitro wound healing was used to evaluate cell migration. The
pericytes were cultured to near confluence in 24-well plates with or
without precoating with 12.5% extracellular matrix (Matrigel; BD Bio-
sciences) in MEM containing 10% FCS at 37°C. After being maintained
at quiescence in serum-free MEM for 45 minutes, the cell monolayers
were wounded by a 1-cm
3
tip pipette in one direction. The wounded
cells were washed with PBS to remove cellular debris and were
incubated with various concentrations Ang-1 or -2 at 37°C for 7 hours.
Cell migration was monitored at initial wounding and at 7 hours under
a phase-contrast microscope and calculated as migration distance (in
micrometers) ⫽ (distance at time 0 ⫺ distance at time 7 hours)/2.
Visualization of F-Actin
As assembly of actin polymers appears to be an absolute requirement
for pseudopod extension and cell migration, visualization of F-actin
with phalloidin was used as a molecular marker for cell migration.
25
Retinal pericytes grown on slides precoated with extracellular matrix
(Matrigel; BD Biosciences) were treated with 100 ng/mL Ang-1 or -2 at
37°C for 48 hours. The cells were fixed with 4% paraformaldehyde and
permeabilized with 0.1% Triton X-100. Changes in F-actin structures
were detected by incubating the cells for 30 minutes at RT with 0.33
M TRITC-labeled phalloidin, and the cells were examined by fluores-
cence microscopy. A minimum of five cells in five random microscope
fields for each treatment (at least 25 cells) were analyzed.
Statistical Analysis
All experiments were conducted at least three times. Unless otherwise
indicated, all data are expressed as the mean ⫾ SEM. An unpaired,
two-tailed Student’s t-test was used to determine the significance of the
results of Tie-2 expression, proliferation, migration, and apoptosis
assays by analysis of the two group means. The Mann-Whitney test was
used to determine statistical significance in the laser densitometry of
Western blot analysis. Statistical analysis was performed (Minitab, ver.
14; Minitab, Inc., State College, PA), with P ⬍ 0.05 considered statis-
tically significant.
RESULTS
Regulation of Tie-2 Expression in
Retinal Pericytes
We determined the levels of both the cellular and soluble forms
of Tie-2. (A soluble form of Tie-2 has been found to be released
into the culture supernatant of other cell types.
26
) Both mem
-
brane and soluble Tie-2 expression was observed in pericytes
grown under standard culture conditions (Fig. 1A). With an-
giopoietin treatment at 100 ng/mL for 48 hours, Tie-2 in the
cell lysate significantly increased by 29.5% (4.03 ng Tie-2/mg of
protein for unstimulated vs. 5.21 ng/mg Ang-1 stimulated) and
11.5% (4.03 ng/mg for unstimulated vs. 4.49 ng/mg Ang-2
stimulated) when stimulated with Ang-1 and -2, respectively. In
the supernatant, however, the soluble Tie-2 significantly de-
creased (by ⬃70%) on angiopoietin stimulation (1.62 ng Tie-
2/mg of protein without stimulation vs. 0.54 ng/mg and 0.44
ng/mg Tie-2 when stimulated with Ang-1 and -2, respectively).
Ang-1 induced tyrosine phosphorylation of Tie-2 in pericytes
within 5 minutes, reaching a maximum level at 1 hour, fol-
lowed by a slight downregulation (Fig. 1B). In contrast, the
kinetics of Tie2 phosphorylation induced by Ang-2 was a grad-
ual upregulation, with the maximum phosphorylation occur-
ring at 48 hours (Fig. 1B).
Flow cytometry demonstrated that more than 98% of cul-
tured pericytes expressed Tie-2 (normal control: 98.4% ⫾
0.5%; 100 ng/mL Ang-1 or -2 for 48 hours: 99.5% ⫾ 0.3% or
99.7% ⫾ 0.2%, respectively; Fig 1C). Compared with normal
control cells, the angiopoietin-treated pericytes exhibited an
increase in levels of Tie-2 shown as MFI (arbitrary units; normal
control: 99.8 ⫾ 17.6; Ang-1: 178.7 ⫾ 9.8; Ang-2: 138.3 ⫾ 14.4;
Fig 1C). Furthermore, the cells exhibited typical pericyte mor-
phology when assessed by microscopy, and greater than 94%
of the cells expressed desmin (see representative flow cyto-
metric analysis in Fig. 1D). Consistent with the desmin expres-
sion, no PECAM-1 (a marker for endothelial cells but not peri-
cytes) was detected in pericyte cultures by Western blot
analysis, whereas microvascular endothelial cells strongly ex-
pressed PECAM-1 (Fig. 1E).
Ang-1 Regulation of CD13 Expression in
Retinal Pericytes
CD13 protein was expressed by retinal pericytes and was
upregulated by Ang-1 in a dose-dependent manner (Fig. 2A).
Tie-2 antisense (but not Tie-2 sense) treatment significantly
reduced CD13 expression (Fig. 2B) and abolished Ang-1-
induced CD13 expression in pericytes (Fig. 2C). Ang-2 had no
effect on CD13 expression in retinal pericytes (data not
shown).
Differential Effect of Ang-1 and -2 on
Pericyte Survival
TNF-
␣
(100 ng/mL) induced maximum apoptosis of pericytes
(25%) compared with baseline (7%) by 24 hours (Fig. 3). Ang-1
at 100 ng/mL significantly inhibited TNF-
␣
-induced apoptosis
with the number of apoptotic cells decreasing by approxi-
mately 50%. By contrast 100 ng/mL Ang-2 enhanced TNF-
␣
-
induced apoptosis reaching up to 30% by 48 hours (Fig. 3B). A
similar response was observed with high glucose, Ang-1 de-
creased glucose-induced apoptosis by 60%, and Ang-2 en-
hanced apoptosis by 47% at 25 mM glucose (Fig 3C).
Effects of Ang-1 and -2 on Retinal
Pericyte Behavior
Proliferation. Retinal pericytes exposed to Ang-2 (be-
tween 1 and 200 ng/mL) exhibited a small but significant
dose-dependent increase in cell proliferation (Fig. 4A),
whereas Ang-1 had no mitogenic effect on the cells.
The effects of Ang-1 and -2 on cell cycle progression were
evaluated by flow cytometric analysis of the DNA content. As
shown in Table 1 most of the cells under control conditions
were in the G
0
/G
1
phase. Exposure to Ang-2 resulted in a
significant (P ⬍ 0.05) shift toward the S-phase and G
2
/M phase,
whereas Ang-1 had no effect compared with control.
Migration. Addition of Ang-1 and -2 had no obvious effect
on retinal pericyte migration on standard cell culture plastic
(Fig. 4B). A recent study
27
showed that an antibody against
CD13 inhibits cell migration on basement membrane matrix
(Matrigel; BD Biosciences). Because in the present study Ang-1
upregulated CD13 expression in the pericytes, we assessed
pericyte migration on basement membrane matrix– coated
slides. As shown in Figure 4B, retinal pericyte migration was
significantly stimulated by Ang-1 in a dose-dependent manner
on basement membrane matrix-coated slides, and the effect
was almost completely inhibited by Tie-2 antisense. There was
no significant increase in pericyte migration after treatment
with Ang-2 (Fig. 4B).
We next determined whether the changes in cell migration
were associated with alterations in the F-actin cytoskeletal
organization. The control cells were flat with thin, uniform,
parallel actin filaments (stress fibers) departing from single
foci and extending throughout the length of the cell, with
the presence of lamellipodia (Figs. 5A,D). There were a few
IOVS, May 2008, Vol. 49, No. 5 Tie-2 in Pericyte Survival and Recruitment 2165
dominant, elongated pseudopodia and a few actin-contain-
ing fine cell extensions. Ang-1 treatment caused dramatic
reorganization of the actin cytoskeleton that resulted in a
reduction in the number of stress fibers and lamellipodia,
with an uneven thickening of the remaining stress fibers and
an increase in both dominant leading-edge pseudopodia
observed at the ends of the cells and in invadopodia (Fig.
5B). By contrast, Ang-2-treated retinal pericytes remained
flat with the partial disassociation of actin into aggregates
(Fig. 5E). Tie-2 antisense treatment abolished the effects of
Ang-1 and -2 on changes in actin cytoskeleton in the retinal
pericytes (Figs. 5C, 5F).
DISCUSSION
Retinal pericytes are intimately associated with the vascular
endothelium and have a critical role in maintaining the func-
tional integrity of the capillary unit.
28
Acting as contractile cells
with a well-developed actin microfilamentous network, peri-
cytes alter capillary luminal diameter and regulate retinal blood
flow.
22
Retinal pericytes also appear to have a critical trophic
function in promoting vessel maturation and endothelial sur-
vival. Selective degeneration of pericytes in the retinal capillary
vessels is a distinguishing feature of early retinal vascular dam-
age in diabetes and can result in decreased capillary tonicity,
formation of microaneurysms, and vessel dilation.
23
It has been proposed that hyperglycemia causes pericyte
dysfunction, apoptosis, and ultimately pericyte loss.
29,30
Our
finding that Ang-2 protein is upregulated in the diabetic retina
is consistent with that observed in a chronic hyperglycemia
animal model of diabetic retinopathy.
2
Ang-2 has different
effects on Tie-2 phosphorylation dependent on the cell types
on which it acts. For example, Ang-2 binds to endothelial cell
Tie-2 without inducing its phosphorylation but can activate
Tie-2 in genetically modified mesenchymal cells.
31
Although
Tie-2 has been presumed to be restricted to endothelial cells
and hematopoietic stem cells, we provide evidence that Tie-2
receptors are expressed by retinal pericytes, confirming other
reports that Tie-2 expression can be found in nonendothelial
mesenchymal cell types.
10,11,32
Tie-2 immunolocalization to
nonvascular cells in both the inner and outer retina is in
agreement with that reported by Ohashi et al.
33
FIGURE 1. Tie-2 expression in reti-
nal pericytes. (A) Lysates and super-
natants from retinal pericyte cultures
were analyzed for Tie-2 by ELISA.
Tie-2 levels were normalized for total
protein concentrations in the cell ly-
sates. Tie-2 was present in the lysate
of unstimulated pericytes and in-
creased on addition of Ang-1 and -2
for 48 hours. The levels of soluble
Tie-2 in the medium decreased for
pericytes treated with Ang-1 and -2.
(B) The cell lysates were immuno-
precipitated with anti-tyrosine phos-
phorylation antibody (PY20). The im-
munoprecipitated proteins were
subject to ELISA analysis of Tie-2. The
results are represented as the
mean ⫾ SEM *P ⬍ 0.05 versus the
control group. To confirm the speci-
ficity of PY20 immunoprecipitation,
cell lysates immunoprecipitated with
PY20 and subjected to Western blot
analysis demonstrated a single band
for Tie-2 (inset). (C) The pericytes
were analyzed by flow cytometry for
expression of Tie-2. Top left: percent-
ages of pericytes bearing Tie-2 im-
mune reactivity at the pericyte sur-
face and mean fluorescence intensity
(MFI) of Tie-2-positive cells. (D) The
percentage of cells within the cul-
ture staining desmin-positive is
shown, to confirm the purity of the
pericyte cultures with increasing pas-
sages. Data represent the mean re-
sults of three experiments. Vertical
bars: upper limit of the negative iso-
type control. (E) Microvascular endo-
thelial cell contamination of pericyte
cell lysates was ruled out by Western
blot analysis with an anti-PECAM-1
antibody (microvascular endothelial
cell lysates were used as the positive
control). Data are expressed as the
mean ⫾ SEM (n ⫽ 3).
2166 Cai et al. IOVS, May 2008, Vol. 49, No. 5
Both Ang-1 and -2 enhanced Tie-2 expression in cultured
retinal pericytes, indicating that upregulation rather than con-
stitutive expression of Tie-2 occurs during the progression of
diabetic retinopathy. Consistent with these findings, both
Ang-1 and -2 induced Tie-2 autophosphorylation, although ro-
bust Tie-2 activation by Ang-2 required prolonged exposure to
Ang-2. Furthermore, the decrease in sTie-2 levels in pericyte-
conditioned media after Ang-1 or -2 treatment suggests that
Tie-2 may not be susceptible to proteolysis after ligand binding.
34
Retinal capillary pericytes extend long cytoplasmic pro-
cesses to form interdigitating contacts with endothelial cells
that facilitate the maturation, remodeling, and maintenance of
the vascular system. During retinal blood vessel development,
Ang-2 acts as a factor that primes endothelial cells for angio-
genesis by destabilizing interactions between endothelial cells
and perivascular cells.
18
Increased Ang-2 is thought to lead to
persistent disruption of the cellular cross-talk between peri-
cytes and endothelial cells in early diabetic retinopathy, culmi-
nating in pericyte loss and vessel destabilization. However, our
data that Ang-2 promotes retinal pericyte proliferation in a
dose-dependent manner add a new aspect to the complexity of
the Ang-2/Tie-2 system. A recent report
35
suggests that mitot
-
ically active pericytes can form angiogenic sprouts or tubes
during the early phase of neovascularization in tumors and in
developing retinal tissues. Although Ang-1 has been identified
as the primary activating ligand for Tie-2, Ang-2 possesses
similar receptor affinity to Tie-2 and it has been proposed that
in response to Ang-1, Tie-2 is rapidly internalized and targeted
FIGURE 2. Ang-1 upregulated CD13
(aminopeptidase N) expression in
retinal pericytes. Western blots are
from a representative experiment.
The densitometric analysis incorpo-
rates the mean results of at least
three separate experiments. (A)
Whole-cell lysates from the retinal
pericytes treated with 100 ng/mL
Ang-1 for 48 hours were probed with
an antibody for CD13. Ang-1 treat-
ment led to a dose-dependent in-
crease in CD13 (150 kDa) expression
in retinal pericytes. Densitometric
analyses are presented as the relative
ratio of CD13 to
␣
-tubulin (55 kDa)
and the ratio relative to control is
arbitrarily presented as 1. Data are
expressed as the mean ⫾ SEM (n ⫽
4). *P ⬍ 0.05, **P ⬍ 0.01 versus the
control. (B) The pericytes exposed
to Tie-2 antisense eliminated the Ang-
1-induced expression of CD13,
whereas Tie-2 sense had no effect.
Densitometric analyses are presented
as the relative ratio of CD13 to
␣
-tu-
bulin, and the ratio relative to control
is arbitrarily presented as 1. Data are
expressed as the mean ⫾ SEM (n ⫽
4).**P ⬍ 0.001 versus Ang-1 only. (C)
Oligo at 6
M was the optimal con-
centration for antisense Tie-2, as de-
termined by Western blot analysis of
Tie-2 (140 kDa) expression by peri-
cytes after transfection with different
concentrations of the antisense Tie-2
oligos. Densitometric analyses are
presented as the relative ratio of
Tie-2 to
␣
-tubulin, and the ratio rela-
tive to control is arbitrarily presented
as 1. Data are expressed as the
mean ⫾ SEM (n ⫽ 4). **P ⬍ 0.01
versus 0
M Tie-2 antisense.
IOVS, May 2008, Vol. 49, No. 5 Tie-2 in Pericyte Survival and Recruitment 2167
for degradation.
36
By contrast, Ang-2 only weakly activates
Tie-2 without significantly stimulating Tie-2 internalization,
leading to amplification and a delay before full activation of
Tie-2 occurs.
CD13 was initially identified as an important marker of
subpopulations of hematopoietic cells.
20
CD13 is identical
with a predominant metalloproteinase (MMP), a Zn
2⫹
-depen
-
dent ectopeptidase
37
that activates or inactivates bioactive
peptides on the cell surface by preferential cleaving of proteins
with NH
2
-terminal neutral amino acids
38,39
and that regulates
the availability of peptides to adjacent cells. A recent study
showed that CD13 is expressed by vascular cells and may play
a role in angiogenesis.
40
In our study, retinal pericytes also
expressed CD13 and Ang-1 enhanced retinal pericyte CD13
expression in a dose-dependent manner. Recently, MMPs have
been reported to be involved in tumor pericyte recruitment.
41
The strong enhancement of a migratory cell phenotype (an
increase in actin polymerization and pseudopod extension) in
Ang-1-treated pericytes on synthetic matrix (Matrigel; BD Bio-
sciences) in this study, accompanied with increased expres-
sion CD13, indicates that Ang-1 indeed increased the motility
of pericytes, at least in part, via CD13 mediated-MMP activity.
Moreover, that Ang-1 stimulated pericyte migration but not
proliferation reinforces our hypothesis that interaction of
Ang-1 with Tie-2 in pericytes leads to vessel maturation via
recruitment of pericytes and smooth muscle cells. Of interest,
CD13 has also been reported to be a putative marker for
cerebral pericyte maturation.
20
It has been proposed that overexpression of Ang-1 in vivo
results in a dramatic increase in microvessel number and vessel
FIGURE 3. The effect of angiopoi-
etins on TNF-
␣
-induced apoptosis of
retinal pericytes. Apoptosis was in-
duced in TNF-
␣
-treated retinal peri-
cytes. Pericytes, exposed to Ang-1
and -2 (100 ng/mL), either in the
presence or absence of TNF-
␣
(100
ng/mL) for the indicated times, were
stained with annexin-V-FITC and pro-
pidium iodide and analyzed by flow
cytometry. (A) Representative exper-
iment from flow cytometric analysis
of annexin-V FITC/propidium iodide
stained pericytes. (B) Cell apoptosis
was expressed as the percentage of
apoptotic cells in the total cell pop-
ulation. TNF-
␣
-induced maximum
apoptosis of pericytes by 24 hours.
Ang-2 enhanced TNF-
␣
-induced apo-
ptosis with maximum effect at 48
hours, whereas Ang-1 significantly re-
duced apoptosis. Data are expressed
as the mean ⫾ SEM of results in three
separate experiments. (C) Apoptosis
in pericytes exposed to high glucose
(15 and 25 mM) and the effects of
Ang-1 and -2 as a percentage of apo-
ptotic cells in the total cell popula-
tion. Data are expressed as the
mean ⫾ SEM of results in three sep-
arate experiments.
2168 Cai et al. IOVS, May 2008, Vol. 49, No. 5
branching
42
indicating that Ang-1 plays an important role in
retinal neovascularization, including maturation and remodel-
ing, rather than being a effector of diabetic retinopathy. Fur-
thermore, it has been shown that Ang-1 can rescue the vessels
from leakiness caused by VEGF-A without interfering with
induction of angiogenesis.
43
Our observation that retinal peri
-
cytes express Tie-2 raises the possibility that Ang-1 helps to
maintain and stabilize mature vessels by stimulating pericyte
viability. Because Ang-1 also leads to increased Akt activation in
endothelial cells, thus enhancing survival signals,
44
we subse
-
quently tested whether Ang-1 can block retinal pericyte apo-
ptosis induced by high glucose
29,30
or TNF-
␣
(a factor impli
-
cated in the pathogenesis of diabetic retinopathy
45
). Retinal
pericytes treated with Ang-1 demonstrated significantly de-
creased apoptosis compared with those treated with TNF-
␣
or
high glucose alone, indicating that Ang-1 can improve pericyte
survival.
46
As Ang-1 has been shown to be highly expressed in
adult tissue, our results further emphasize its role in maintain-
ing previously developed and mature blood vessels.
It is now becoming evident that the effect of angiopoietins
is highly context dependent and contingent on the nature of
the local environment, the cellular phenotype, and the stage of
the disease. Fiedler and Reiss
47
elegantly demonstrated that
Ang-2 functions are dependent on the presence of other cyto-
kines, and their findings are supported by our observation that
Ang-2 enhanced the apoptotic effect of TNF-
␣
but had no effect
alone. Pericyte phenotype differs as to whether it is in its
quiescent state encased within the vascular basement mem-
brane or proliferating at the vanguard of newly forming blood
vessels. Although we would not expect this phenotype to
proliferate, we would expect it to be susceptible to apoptosis.
However, the angiopoietins would be expected to promote a
different response in the activated pericytes associated with
angiogenesis, and our data support this. Diabetic retinopathy is
a dynamic process and at stages during the disease both peri-
cyte phenotypes will be present. Thus, it is not surprising that
the angiopoietins can have multiple effects dependent on the
context of endogenous and exogenous factors.
In conclusion, our results suggest (1) a critical role for Ang-2
in pericyte loss in early diabetic retinopathy, (2) a possible role
for Ang-2 in pericyte proliferation at a later stage in the pro-
FIGURE 4. Angiopoietins mediate
retinal pericyte proliferation and mi-
gration. (A) Ang-2 modestly in-
creased proliferation of retinal peri-
cytes in a dose-dependent manner,
whereas Ang-1 had no effect. The
number of cells is expressed as the
mean ⫾ SEM of counts in triplicate
wells. *P ⬍ 0.05 versus the unstimu-
lated control group in which Ang
was substituted with 0.1% BSA; (B)
Ang-1 had no effect on the migration
of retinal pericytes cultured on plas-
tic. In contrast, Ang-1 treatment sig-
nificantly enhanced retinal pericyte
migration on a synthetic-matrix–
coated surface. Tie-2 antisense treat-
ment abolished the enhancing effect
of Ang-1 on retinal pericyte migra-
tion on the matrix. However, there
was no significant increase in retinal
pericyte migration after treatment
with Ang-2 in the presence or ab-
sence of matrix. The data are repre-
sented as the mean ⫾ SEM (n ⫽ 3).
T
ABLE 1. Ang-1- and -2-Mediated Pericyte Cell Cycle Progression
Cell Cycle Control Ang-1 Ang-2
G
0
/G
1
88.7 ⫾ 6 87.7 ⫾ 5 67.5 ⫾ 5
S 10.9 ⫾ 2 12.1 ⫾ 3 21.7 ⫾ 3
G
2
/M
0.4 ⫾ 1 0.4 ⫾ 2 11.8 ⫾ 2
Data are the percentage of cells ⫾ SEM.
IOVS, May 2008, Vol. 49, No. 5 Tie-2 in Pericyte Survival and Recruitment 2169
gression of diabetic retinopathy, and (3) Ang-1-mediated im-
provement of survival, activation, and migration of retinal peri-
cytes during the establishment of retinal new vessels. Targeted
regulation of angiopoietin isoforms may offer a therapeutic
approach as an adjunct treatment for the microvascular com-
plications associated with diabetic retinopathy. Das et al.
48
have reported that a Tie-2 antagonist, muTek delta Fc, can
inhibit revascularization in an oxygen model of retinopathy.
References
1. Miller J, D’Amico D. Proliferative diabetic retinopathy. In: Albert D,
Jakobiec F, eds. Principles and Practice of Ophthalmology.
Philadelphia: WB Saunders; 1994:760 –782.
2. Hammes HP, Lin J, Wagner P, et al. Angiopoietin-2 causes pericyte
dropout in the normal retina: evidence for involvement in diabetic
retinopathy. Diabetes. 2004;53:1104 –1110.
3. Cai J, Boulton M. The pathogenesis of diabetic retinopathy: old
concepts and new questions. Eye. 2002;16:242–260.
4. Ishibashi T, Inomata H, Sakamoto T, Ryan SJ. Pericytes of newly
formed vessels in experimental subretinal neovascularization. Arch
Ophthalmol. 1995;113:227–231.
5. Takagi H, Koyama S, Seike H, et al. Potential role of the angiopoi-
etin/tie2 system in ischemia-induced retinal neovascularization.
Invest Ophthalmol Vis Sci. 2003;44:393– 402.
6. Jones N, Master Z, Jones J, et al. Identification of Tek/Tie2 binding
partners: binding to a multifunctional docking site mediates cell
survival and migration. J Biol Chem. 1999;274:30896 –30905.
7. Jones N, Voskas D, Master Z, Sarao R, Jones J, Dumont DJ. Rescue
of the early vascular defects in Tek/Tie2 null mice reveals an
essential survival function. EMBO Rep. 2001;2:438 – 445.
8. Fadel BM, Boutet SC, Quertermous T. Octamer-dependent in vivo
expression of the endothelial cell-specific Tie2 gene. J Biol Chem.
1999;274:20376–20383.
9. Maisonpierre PC, Suri C, Jones PF, et al. Angiopoietin-2, a natural
antagonist for Tie2 that disrupts in vivo angiogenesis. Science.
1997;277:55–60.
10. Davis S, Aldrich TH, Jones PF, et al. Isolation of angiopoietin-1, a
ligand for the Tie2 receptor, by secretion-trap expression cloning.
Cell. 1996;87:1161–1169.
11. Puri MC, Bernstein A. Requirement for the TIE family of receptor
tyrosine kinases in adult but not fetal hematopoiesis. Proc Natl
Acad Sci USA. 2003;100:12753–12758.
12. Park YS, Kim NH, Jo I. Hypoxia and vascular endothelial growth
factor acutely up-regulate angiopoietin-1 and Tie2 mRNA in bovine
retinal pericytes. Microvasc Res. 2003;65:125–131.
13. Suri C, Jones PF, Patan S, et al. Requisite role of angiopoietin-1, a
ligand for the TIE2 receptor, during embryonic angiogenesis. Cell.
1996;87:1171–1180.
14. Hawighorst T, Skobe M, Streit M, et al. Activation of the tie2
receptor by angiopoietin-1 enhances tumor vessel maturation and
impairs squamous cell carcinoma growth. Am J Pathol. 2002;160:
1381–1392.
15. Kampfer H, Pfeilschifter J, Frank S. Expressional regulation of
angiopoietin-1 and -2 and the Tie-1 and -2 receptor tyrosine kinases
during cutaneous wound healing: a comparative study of normal
and impaired repair. Lab Invest. 2001;81:361–373.
16. Yamakawa M, Liu LX, Belanger AJ, et al. Expression of angiopoi-
etins in renal epithelial and clear cell carcinoma cells: regulation
by hypoxia and participation in angiogenesis. Am J Physiol. 2004;
287:F649–F657.
17. Hanahan D. Signaling, vascular morphogenesis and maintenance.
Science. 1997;277:48 –50.
18. Hackett SF, Ozaki H, Strauss RW, et al. Angiopoietin 2 expression
in the retina: upregulation during physiologic and pathologic neo-
vascularization. J Cell Physiol. 200;184:275–284.
19. Cai J, Ahmad S, Jiang WG, et al. Activation of vascular endothelial
growth factor receptor-1 sustains angiogenesis and Bcl-2 expres-
sion via the phosphatidylinositol 3-kinase pathway in endothelial
cells. Diabetes. 2003;52:2959 –2968.
20. Griffin JD, Ritz J, Nadler LM, Schlossman SF. Expression of myeloid
differentiation antigens on normal and malignant myeloid cells.
J Lab Invest. 1981;68:932–941.
21. Ramsauer M, Krause D, Dermietzel R. Angiogenesis of the blood-
brain barrier in vitro and the function of cerebral pericytes. FASEB
J. 2002;16:1274 –1276.
22. Cai J, Jiang WG, Grant MB, Boulton M. Pigment epithelium-derived
factor inhibits angiogenesis via regulated intracellular proteolysis
of vascular endothelial growth factor receptor 1. J Biol Chem.
2006;281:3604–3613.
23. Lauffenburger DA, Horwitz AF. Cell migration: a physically inte-
grated molecular process. Cell. 1996;84:359 –369.
24. Reusch P, Barleon B, Weindel K, et al. Identification of a soluble
form of the angiopoietin receptor TIE-2 released from endothelial
cells and present in human blood. Angiogenesis. 2001;4:123–131.
25. Ikeda N, Nakajima Y, Tokuhara T, et al. Clinical significance of
aminopeptidase N/CD13 expression in human pancreatic carci-
noma. Clin Cancer Res. 2003;9:1503–1508.
FIGURE 5. Cytoskeletal changes in retinal pericytes exposed to Ang-1
and -2. Retinal pericytes were fixed and stained with phalloidin-TRITC.
A minimum of five cells in five random microscope fields for each
treatment (a total of at least 25 cells) were analyzed. Control cells
showed that thin, uniform, parallel actin filaments with few elongated
pseudopodia (arrowhead)(A, D). Treatment with 100 ng/mL Ang-1
resulted in a reduction in the number of stress fibers, with an uneven
thickening of the remaining fibers in the dominant leading-edge pseu-
dopodia observed at the ends of the cells (arrow)(B). Ang-2 (100
ng/mL)-treated pericytes remained flat with partial dissociation of actin
into aggregates (chevron)(E). Tie-2 antisense treatment abolished the
effects of both Ang-1 and -2 on changes in actin cytoskeleton in the
retinal pericytes (C, F). Bar, 50
m.
2170 Cai et al. IOVS, May 2008, Vol. 49, No. 5
26. Kern TS, Engerman RL. Pharmacological inhibition of diabetic
retinopathy: aminoguanidine and aspirin. Diabetes. 2001;50:
1636–1642.
27. Chakravarthy U, Gardiner TA. Endothelium-derived agents in peri-
cyte function/dysfunction. Prog Retin Eye Res. 1999;18:511–527.
28. Lindahl P, Johansson BR, Leveen P, Betsholtz C. Pericyte loss and
microaneurysm formation in PDGF-B-deficient mice. Science.
1997;277:242–245.
29. Miller AG, Smith DG, Bhat M, Nagaraj RH. Glyoxalase is critical for
human retinal capillary pericyte survival under hyperglycaemic
conditions. J Biol Chem. 2006;281:11864 –11871.
30. Romeo G, Liu WH, Asnaghi V, Kern TS, Lorenz M. Activation of
nuclear factor-
B induced by diabetes and high glucose regulates a
proapoptotic program in retinal pericytes. Diabetes. 2002;51:
2241–2248.
31. Teichert-Kuliszewska K, Maisonpierre PC, Jones N, et al. Biological
action of angiopoietin-2 in a fibrin matrix model of angiogenesis is
associated with activation of Tie2. Cardiovasc Res. 2001;49:659 –
670.
32. Iurlaro M, Scatena M, Zhu WH, et al. Rat aorta-derived mural
precursor cells express the Tie2 receptor and respond directly to
stimulation by angiopoietins. J Cell Sci. 2003;116:3635–3643.
33. Ohashi H, Takagi H, Koyama S, et al. Alterations in expression of
angiopoietins and the Tie-2 receptor in the retina of streptozotocin
induced diabetic rats. Mol Vis. 2004;10:608 – 617.
34. Yabkowitz R, Meyer S, Black T, Elliott G, Merewether LA, Yamane
HK. Inflammatory cytokines and vascular endothelial growth fac-
tor stimulate the release of soluble tie receptor from human en-
dothelial cells via metalloprotease activation. Blood. 1999;93:
1969–1979.
35. Ozerdem U, Stallcup WB. Early contribution of pericytes to angio-
genic sprouting and tube formation. Angiogenesis. 2003;6:241–
249.
36. Bogdanovic E, Nguyen VPKH, Dumont DJ. Activation of Tie2 by
angiopoietin-1 and angiopoietin-2 results in their release and re-
ceptor internalization. J Cell Sci. 2006;119:3551–3560.
37. Shimamura M, Hazato T, Iwaguchi T. Enkephalin-degrading ami-
nopeptidase in the longitudinal muscle layer of guinea pig small
intestine: its properties and action on neuropeptides. J Biochem.
1991;109:492–497.
38. Olsen J, Cowell GM, Konigshofer E, et al. Complete amino acid
sequence of human intestinal aminopeptidase N as deduced from
cloned cDNA. FEBS Lett. 1988;238:307–314.
39. Ashmun RA, Look AT. Metalloprotease activity of CD13/aminopep-
tidase N on the surface of human myeloid cells. Blood. 1990;75:
462–469.
40. Pasqualini R, Koivunen E, Kain R, et al. Aminopeptidase N is a
receptor for tumor-homing peptides and a target for inhibiting
angiogenesis. Cancer Res. 2000;60:722–727.
41. Chantrain CF, Henriet P, Jodele S, et al. Mechanisms of pericyte
recruitment in tumour angiogenesis: a new role for metallopro-
teinases. Eur J Cancer. 2006;42:310 –318.
42. Machein MR, Knedla A, Knoth R, Wagner S, Neuschl E, Plate KH.
Angiopoietin-1 promotes tumor angiogenesis in a rat glioma
model. Am J Pathol. 2004;165:1557–1570.
43. Gamble JR, Drew J, Trezise L, et al. Angiopoietin-1 is an antiper-
meability and anti-inflammatory agent in vitro and targets cell
junctions. Circ Res. 2000;87:603– 607.
44. Kim I, Kim H G, So JN, Kim JH, Kwak HJ, Koh GY. Angiopoietin-1
regulates endothelial cell survival through the phosphatidylinositol
3⬘-kinase/Akt signal transduction pathway. Circ Res. 2000;86:24 –
29.
45. Gardiner TA, Gibson DS, de Gooyer TE, de la Cruz VF, McDonald
DM, Stitt AW. Inhibition of tumor necrosis factor-alpha improves
physiological angiogenesis and reduces pathological neovascular-
ization in ischemic retinopathy. Am J Pathol. 2005;166:637– 644.
46. Joussen AM, Poulaki V, Mitsiades N. Nonsteroidal anti-inflamma-
tory drugs prevent early diabetic retinopathy via TNF-alpha sup-
pression. FASEB J. 2002;16:438 – 440.
47. Fiedler U, Reiss Y, Scharpfenecker M, et al. Angiopoietin-2 sensi-
tizes endothelial cells to TNF-alpha and has a crucial role in the
induction of inflammation. Nat Med. 2006;12:235–239.
48. Das A, Fanslow W, Cerretti D, et al. Angiopoietin/Tek interactions
regulate MMP-9 expression and retinal neovascularization. Lab
Invest. 2003;83:1637–1645.
IOVS, May 2008, Vol. 49, No. 5 Tie-2 in Pericyte Survival and Recruitment 2171