Cerebral Capillary Endothelial Cell Mitogenesis and
Morphogenesis Induced by Astrocytic
Epoxyeicosatrienoic Acid
Chenyang Zhang, MD, PhD; David R. Harder, PhD
Downloaded from http://stroke.ahajournals.org/ by guest on June 18, 2017
Background and Purpose—Epoxyeicosatrienoic acids (EETs) are products of cytochrome P450 epoxygenation of
arachidonic acid. We have previously demonstrated that astrocyte-conditioned medium induced mitogenesis in brain
capillary endothelial cells. The goals of the present studies are to further define the mechanism through which this can
occur and to confirm that EETs are derived from astrocytes, through which astrocytic activity can regulate cerebral
angiogenesis in response to neuronal activation.
Methods—Astrocytes and cerebral capillary endothelial cells in primary cultures were cocultured to examine the
interaction of the 2 cell types. We used multiple immunohistochemical techniques to characterize the multicellular
nature of the capillaries, which is not simply an artifact related to the culture conditions. The mitogenic effect of EETs
was determined by 3H-thymidine incorporation and cell proliferation assay. Endothelial tube formation was examined
in vitro and in vivo with the use of a reconstituted basement membrane (Matrigel) assay.
Results—In cocultures of astrocytes and capillary endothelium, we observed morphological changes in both cell types such
that each assumed certain physiological characteristics, ie, endothelial networks and astrocytes with “footlike”
projections as well as intermittent gap junctions forming within the endothelial cells. EETs from astrocytes as well as
synthetic EETs promoted mitogenesis of endothelial cells, a process sensitive to inhibition of tyrosine kinase with
genistein. Treatments with exogenous EETs were sufficient for endothelial cells to differentiate into capillary-like
structures in culture as well as in vivo in a Matrigel matrix.
Conclusions—The 2 major conclusions from these data are that astrocytes may play an important role in regulating
angiogenesis in the brain and that cytochrome P450 – derived EETs from astrocytes are mitogenic and angiogenic.
(Stroke. 2002;33:●●●-●●●.)
Key Words: angiogenesis 䡲 capillaries 䡲 cells, cultured 䡲 cytochrome P-450 䡲 gap junctions 䡲 rats
A
strocytes in the central nervous system are anatomically
situated between neurons and the microcirculation,
forming “foot processes” that impinge on arteriolar and
capillary networks. Traditionally, it has been assumed that
astrocytes play an organizational role, keeping neurons and
capillary networks in register and supporting the blood-brain
barrier. Recently, the roles of astrocytes with respect to cross
talk with neurons and cerebral vasculature have been reexamined, and it has been suggested that astrocytes could be
key regulatory elements in the brain.1–3 In the retina, astrocytes have been reported to be required for capillary angiogenesis since the migration of astrocytes into developing
retina precedes formation of the retinal vasculature.4 –7 The
mechanism underlying astrocyte-involved angiogenesis in the
brain is largely unknown.
In a previous communication8 we showed that astrocyteconditioned culture media stimulated proliferation of cerebral
microvascular endothelial cells in culture. Formation of
capillary tubes in the coculture of astrocytes and endothelial
cells was blocked on inhibition of cytochrome P450 (P450)
enzymes.8 One such enzyme is encoded by the cytochrome
P450 2C11 (CYP2C11) gene. We have cloned and sequenced
CYP2C11 cDNA from astrocytes of rats.9 Epoxyeicosatrienoic acids (EETs) are biologically active metabolites of
arachidonic acid (AA), the formation of which is catalyzed by
P450 epoxygenase. There are 4 regioisoforms of EETs,
namely, 5,6-EETs, 8,9-EETs, 11,12-EETs, and 14,15EETs.10,11 Previous studies have shown synthesis of EETs by
astrocytes.9,12 Under normal conditions, EETs were found in
the cerebrospinal fluid measured in micromolar concentrations.3 Pharmacological inhibition of EET formation blocked
functional hyperemic response to glutamate infusion in the
brain, suggesting a role of EETs in regulating cerebral blood
flow to match neuronal activity.14 –16 Therefore, we hypothesized that P450-derived EETs were involved in astrocyteinduced mitogenesis and morphogenesis of cerebral capillary
Received April 9, 2002; final revision received June 21, 2002; accepted July 2, 2002.
From the Cardiovascular Research Center, Department of Physiology, Medical College of Wisconsin, Milwaukee.
Correspondence to David R. Harder, PhD, Cardiovascular Research Center, Medical College of Wisconsin, Milwaukee, WI 53226. E-mail
[email protected]
© 2002 American Heart Association, Inc.
Stroke is available at http://www.strokeaha.org
DOI: 10.1161/01.STR.0000037787.07479.9A
1
2
Stroke
December 2002
Downloaded from http://stroke.ahajournals.org/ by guest on June 18, 2017
Figure 1. Cerebral capillary endothelial cell
culture preparation and characterization.
Flowchart describing detailed steps in preparing the cerebral capillary endothelial
cells is shown on the left. a, Phasecontrast image of vascular tissue isolated
from cerebral tissue homogenate consisting mainly of capillary segments. b, Cerebral capillary endothelial cells grown in
L-valine–free medium at culture day 9. c,
Cerebral capillary endothelial cells grown
in regular L-valine medium at culture day 9.
Note that endothelial cells have been overrun with nonendothelial cells (arrows) in
regular L-valine– containing medium, indicating that L-valine–free medium helped to
gain pure endothelial cells in culture. d,
Cerebral capillary endothelial cells stained
positive for DiI-Ac-LDL. e, Confluent cerebral capillary endothelial cells in culture
(nuclei stained with DAPI) were immunoreactive to PECAM-1 at cell membrane, a
specific marker for endothelial cells.
Bar⫽0.1 mm for a through d; bar⫽30 ␮m
for e.
endothelial cells. The role of EETs as intercellular and
intracellular mediators involved in cell proliferation and
angiogenesis in the central nervous system has never been
systematically defined before the present study.
In this report we define the actions of P450-derived EETs,
exogenously added or released from astrocytes, on proliferation and differentiation of cerebral capillary endothelial
cells. We demonstrate that EETs function as angiogenic
regulators for astrocyte-mediated capillary network
formation.
Materials and Methods
Cell Cultures
We prepared primary cultures of astrocytes from hippocampi of
postnatal 3-day-old rat pups as described previously.8,9 Cerebral
capillary endothelial cells were prepared from brains of 4-week-old
rats as described previously8 with some modifications, as outlined in
Figure 1. Briefly, the cerebral cortex was dissected and homogenized
in ice-cold HEPES-buffered salt solution with 0.9% glucose. Vascular tissues were separated from the rest of brain tissue by
centrifugation in a HEPES-buffered salt solution containing 15%
dextran. The vascular tissue was filter through a 150-mesh screen to
remove large vessels. The elutes were loaded on a glass bead
column. Capillaries adhering to the beads were released by sharply
shaking the beads in buffer. The microvessel pellet was digested with
collagenase (500 ␮g/mL) in RPMI-1640 (Biowhitaker) containing
10% fetal bovine serum (FBS) for 15 minutes at room temperature.
The formation of single cell suspension was monitored under a
phase-contrast microscope. After centrifugation, the cell pellet was
resuspended in a L-valine–free medium (Life Technologies), which
inhibits growth of other cell types but not endothelial cells,16,17 and
plated in T25 flasks precoated with fibronectin at 5 ␮g/cm2. Cells
were incubated at 37°C in a 95%/5% mixture of atmospheric air and
CO2. After 3 days, the medium was changed to microvascular
endothelial cell growth (MV) medium formulated to promote endothelial cell growth. This medium is made from serum-free endothelial cell basal medium-2 (EBM-2) supplemented with vascular
endothelial growth factor (VEGF), epidermal growth factor, fibroblast growth factor, insulin-like growth factor, ascorbic acid, hydrocortisone, heparin, FBS, and antibiotics, according to the manufacturer’s instructions (Clonetics). Confluent first passage of endothelial
cells was used for our experiments.
Coculture of Astrocytes and Cerebral Capillary
Endothelial Cells
To separate microglial cells from astrocytes, cultures were shaken
for 60 minutes (at 225 rpm) at 37°C on an orbital shaker.18 The
adherent astrocyte monolayer was reseeded. Cocultures of astrocytes
and endothelial cells were made as described previously.19 Briefly,
confluent astrocytes were trypsinized and resuspended in Dulbecco’s
modified Eagle’s medium containing 10% FBS and antibiotics. Cells
were plated at 6000 cells per square centimeter on fibronectin-coated
coverslips. After 24 hours of incubation, the medium was removed,
and endothelial cells were plated at approximately 20 000 cells per
coverslip in MV medium. Control coverslips lacking either the
astrocytes or endothelial cells were incubated in the same medium as
used for cocultures. After another 24 hours, some cultures were
changed to EBM-2 containing 0.1% bovine serum albumin (BSA).
Medium was changed every 3 days until processed for further
experiments.
Treatments
For preparation of conditioned medium from astrocytes, subconfluent astrocyte cultures were washed with Dulbecco’s PBS, and
Dulbecco’s modified Eagle’s medium with 0.1% BSA (fatty acid
free) was added to cells for conditioning overnight. To inhibit P450
activity in astrocytes, 17-octadecynoic acid (17-ODYA) at 10
␮mol/L was present during the whole time of conditioning. Other
treatments for endothelial cells used were at the following concentrations: VEGF 1 nmol/L, Myr␺PKC-I 100 nmol/L, and genistein
100 ␮mol/L. Vehicle, either ethanol or buffer in the same volume as
treatments, was used as control.
3
H-Thymidine Incorporation
Confluent cerebral endothelial cells in T25 flasks were detached by
trypsin. After centrifugation, the cells were suspended in MV
medium and plated at approximately 7500 cells per well into 24-well
plates and incubated for 1 day so that they reached 80% confluence.
The medium was then changed to EBM-2 medium with 0.1% BSA
for 2 days to make cells quiescent. EETs at various concentrations in
the presence or absence of inhibitors or VEGF were added to the
medium and incubated for 18 hours. 3H-Thymidine at a concentration of 2 ␮Ci/mL was then added to pulse the cells for an additional
3 hours. Cells were washed 3 times with PBS and precipitated with
ice-cold 15% trichloroacetic acid for 30 minutes at 4°C. Wells were
washed gently with water and allowed to try. Cells were lysed with
1N NaOH and incubated at 37°C for 30 minutes. After neutralization
with 1N HCl, the radioactivity from the sample of each well was
Zhang and Harder
EETs in Endothelial Cell Morphogenesis
3
Figure 2. Astrocytes in culture immunolabeled for different cell type markers
and counterstained with DAPI. a, Astrocytes in culture immunolabeled with antibodies against GFAP and counterstained
with DAPI show that most cells (identified by positive DAPI staining in nucleus)
are astrocytes. In confluent astrocyte
cultures, contamination in a few macrophages/microglial cells was identified by immunohistochemistry for OX-42 (arrowheads in b), and no detectable neuron contamination
was identified by immunohistochemistry for neurofilament 200 (NF200) (c). Some nonspecific nuclear staining by neurofilament 200 was
noted. Bar⫽60 ␮m.
determined by liquid scintillation spectrometry. Results were expressed as counts per minute per well. Each experimental data point
represented quadruplicate wells from at least 3 independent experiments. All values were expressed as mean⫾SD. Paired Student’s t
tests were used to compare vehicle control and treatment. P⬍0.05
was considered significant.
Immunocytochemistry
Downloaded from http://stroke.ahajournals.org/ by guest on June 18, 2017
Cells on coverslips or brain sections of 15-␮m thickness were fixed
with 4% paraformaldehyde. After they were blocked with 2% BSA
in PBS, cells or sections were incubated overnight at 4°C with
primary antibodies for single or double labeling. Specific primary
antibodies used were platelet endothelial cell adhesion molecule-1
(PECAM-1) (1:1000, gift from Dr P. Newman), glial fibrillary acidic
protein (GFAP) (1:150, Chemicon), OX-42 (1:100, Serotec), neurofilament 200 (1:200, Sigma), and NeuN (1:100, Chemicon). Secondary antibodies conjugated with FITC and/or TRITC (1:150, Chemicon) were incubated subsequently for 1 hour at room temperature
under dark. For connexin 43 (Santa Cruz), coverslips were incubated
with connexin 43 antibodies at 1:10 overnight, followed by rabbit
anti-goat IgG conjugated to horseradish peroxidase at 1:200 for 4
hours, and labeling was detected by the diaminobenzidine reaction.
After nuclei were counterstained with 4⬘,6-diamidino-2phenylindole (DAPI), sections or cells were rinsed, and coverslips
were placed. Images were taken with Nikon E600 equipped with
epifluorescence and a digital camera. No signal was detected on
control cells or sections, which were processed in the same way
except that primary antibodies were omitted.
DiI-Ac-LDL Staining
After they were washed with PBS, cells were incubated with
DiI-Ac-LDL (Biomedical Technologies Inc) at a concentration of 10
␮g/mL in serum-free EBM-2 for 2 hours at 37°C. Cells were then
washed and fixed with 4% paraformaldehyde for 15 minutes,
followed by placement on coverslips or continuation of the immunocytochemistry protocol. The staining of DiI-Ac-LDL was examined under an inverted microscope.
Morphogenesis Assay on Matrigel
Briefly, endothelial cells were trypsinized and resuspended in MV
medium to inactivate trypsin. After centrifugation, medium was
removed, and cells were resuspended in plain EBM-2, plated at
2⫻104 cells per well into 4-well chamber slides coated with thin
reconstituted basement membrane (Matrigel, Becton Dickinson), and
incubated in the presence or absence of EETs (100 nmol/L or 200
nmol/L for 8,9-EET; 200 nmol/L for 11,12-EET). Some wells were
incubated with MV medium. Eight to 18 hours later, morphology of
endothelial cells was examined under an inverted phase-contrast
microscope, and images were taken with an attached camera. The
relative lengths of tube formed in control or treated conditions were
measured on multiple images that covered the whole area of each
well. The increase in tube formation was evaluated by the ratio of the
total lengths of tubes in treated wells to those of controls. Paired
Student’s t tests were used to compare vehicle control and treatment.
P⬍0.05 was considered significant. The result was reproducible in at
least 3 independent experiments.
Matrigel Implantation In Vivo
Matrigel implantation was conducted as described in detail previously.20 While being anesthetized with halothane, adult C57BL mice
or Sprague-Dawley rats (aged 7 weeks; n⫽5 per group) were each
injected subcutaneously near the abdominal midline with 0.5 mL
Matrigel supplemented with or without 200 nmol/L 8,9- or 11,12EETs with a 25-gauge needle. The animals were kept under
halothane until the injected Matrigel rapidly formed a single, solid
gel. After 7 to 9 days, animals were killed, and gels were recovered
and fixed immediately after dissection in 4% paraformaldehyde.
Matrigel blocks were dehydrated and embedded in paraffin. Paraffin
sections at 5-␮m thickness were stained with hematoxylin and eosin.
The number of vascular structures containing red blood cells per
⫻40 microscopic view on each section was counted and expressed as
mean⫾SD. Paired Student’s t tests were used to compare vehicle
control and treatment. P⬍0.05 was considered significant. The result
was reproducible in 3 independent experiments.
Results
Interaction of Astrocytes and Capillary
Endothelial Cells in Coculture
As shown in Figure 1, we were able to isolate the capillary
fraction (Figure 1a) from the cerebral cortical tissue homogenates. Endothelial cells in culture exhibited classic cobblestone or spindle-shaped morphology. The use of L-valine–free
medium combined with specially formulated microvascular
endothelial cell growth medium helped to ensure high purity
of endothelial cells and characteristic morphology (Figure 1b,
1d, 1e) compared with results obtained with the use of RPMI
1640 with 10% FBS, which contained contaminated cell
types (arrows in Figure 1c). The contaminated cells often
increased in number with time and overran the endothelial
cells. The confluent cerebral capillary endothelial cells were
positive for DiI-Ac-LDL (Figure 1d) and immunoreactive to
PECAM-1 (Figure 1e), an adhesive molecule used as a
marker for endothelial cells. We used endothelial cell cultures
with 93⫾2% purity. The contaminated cell type was mainly
smooth muscle cell, which is not immunoreactive to
PECAM-1 but can be labeled with antibodies against
␣-smooth muscle actin.
Astrocytes in culture exhibited positive immunoreactivity to
GFAP. When becoming confluent, 98⫾1% of cells in culture
were astrocytes identified by GFAP immunoreactivity (Figure 2a). A few macrophage/microglial cells were found in
enriched astrocyte cultures with the use of antibodies against
OX-42 (Figure 2b). Neurons were not detectable in our
confluent astrocyte cultures in which antibodies against
neurofilament 200 (Figure 2c) and NeuN (data not shown)
were used.
4
Stroke
December 2002
Downloaded from http://stroke.ahajournals.org/ by guest on June 18, 2017
Figure 3. Changes in morphology of astrocytes and cerebral capillary endothelial cells when cocultured. a, Formation of capillary-like
structures (arrow) double labeled with DiI-Ac-LDL (red) and GFAP (green) in the coculture of astrocytes and cerebral capillary endothelial cells. b, Formation of capillary-like structures (arrow) triple labeled by PECAM-1 (green), GFAP (red), and DAPI (blue) in the coculture
of astrocytes and cerebral capillary endothelial cells. Note the interaction between astrocytes and endothelial tubes (arrowheads in a
and b). c, Double immunolabeling of blood vessels and astrocytes with PECAM-1 (green) and GFAP (red) in a section from normal rat
cortex showing astrocytes form foot processes that impinge on vessels (arrows). d, Cells (*) forming tube identified by DAPI staining (e)
in coculture exhibit punctate staining of connexin 43 on their cell-cell borders (arrow). Arrow points to area shown in the insert at higher
magnification. f, Cells (*) outside tubes in the same coculture as in d show moderate to light cytoplasmic staining of connexin 43. Bar
in a⫽25 ␮m for a and b; bar⫽20 ␮m in c; bar in f⫽10 ␮m for d, e, and f; bar in the insert of d⫽5 ␮m.
It has been shown previously that astrocytes can induce
formation of endothelial tubelike structures in vitro.8,19 However, the interaction of 2 cell types in coculture has not been
examined in detail. We cocultured primary rat astrocytes and
cerebral capillary endothelial cells over a fibronectin-coated
coverslip. Endothelial cells were identified by either DiI-AcLDL staining or PECAM-1 immunocytochemistry. Astrocytes were detected with antibodies against GFAP. Endothelial cell tubelike structures were evident at day 2, the earliest
time examined. At day 5 in coculture, tubelike networks were
clearly visible under a phase-contrast microscope. As can be
seen in serial microscopic photography, dramatic morphological changes were found in both cell types. Endothelial cells
forming tubes were positive for DiI-Ac-LDL (arrows in
Figure 3a) and PECAM-1 (arrows in Figure 3b). Astrocytes
sent foot processes and impinged on endothelial tubes intermittently (arrowheads in Figure 3a and 3b). Figure 3c shows
a brain section in which endothelial cells were stained with
PECAM-1 and astrocytes were stained with GFAP. A similar
physiological arrangement between astrocytes and endothelial cells in coculture (Figure 3a and 3b) can be seen that
mimics the interaction of astrocytes and endothelial cells in
vivo (Figure 3c). Neither astrocytes nor endothelial cells
showed the aforementioned differentiation when they were
cultured alone. Cells forming tubes (identified by DAPI
staining in Figure 3e) also exhibited punctate staining for
connexin 43 in their cell-cell borders, an evidence of the
formation of gap junction (Figure 3d and insert). However,
cells outside of the tube structures showed light cytoplasmic
staining of connexin 43 (Figure 3f).
EETs Increase 3H-Thymidine Incorporation in
Cerebral Capillary Endothelial Cells
We have shown previously that endothelial tube formation
was inhibited by treating the coculture with 17-ODYA, a
P450 inhibitor. However, we cannot distinguish whether
astrocytic P450 or endothelial P450 was involved. To answer
this question, we treated astrocytes with 17-ODYA and
collected conditioned medium. Astrocyte-conditioned medium in the absence of 17-ODYA showed a significant
mitogenic effect on capillary endothelial cells (Figure 4).
Augmentation in 3H-thymidine incorporation in endothelial
cells was reduced by the conditioned medium that was made
in the presence of 17-ODYA to inhibit P450 enzymatic
activity in astrocytes (Figure 4). Treating endothelial cells
with conditioned medium and 17-ODYA to inhibit endothelial P450 did not block increased 3H-thymidine incorporation
(Figure 4), demonstrating that it was the product of P450 in
astrocyte-conditioned medium, not in endothelial cells, that
was involved in promoting mitogenesis of endothelial cells.
Figure 4. 3H-Thymidine incorporation in primary cerebral capillary endothelial cells (EC) after treatment with astrocyteconditioned media. Conditioned media were prepared as
described in Materials and Methods. Inhibition of P450 activity
by 17-ODYA in astrocytes (AS) significantly (P⬍0.01) attenuated
3
H-thymidine incorporation compared with conditioned medium
without P450 inhibitor. **P⬍0.01 vs vehicle; ##P⬍0.01 vs conditioned media.
Zhang and Harder
EETs in Endothelial Cell Morphogenesis
5
Downloaded from http://stroke.ahajournals.org/ by guest on June 18, 2017
Figure 6. Effects of PKC and tyrosine kinase inhibitors on EETinduced 3H-thymidine incorporation. The tyrosine kinase inhibitor genistein abolished 8,9-EET–stimulated 3H-thymidine incorporation. *P⬍0.05 vs vehicle.
Figure 5. Concentration-dependent stimulation of cerebral capillary endothelial cell proliferation by 8,9-EET. a, Addition of 100
and 200 nmol/L of 8,9-EET in serum-free medium significantly
(P⬍0.001) increased 3H-thymidine incorporation. b, Cell numbers were significantly (P⬍0.001) increased by 8,9-EET and
VEGF treatment. **P⬍0.001 vs vehicle.
The cytochrome P450 2C11 gene encodes an epoxygenase
that produces all 4 regioisomers of EETs in astrocytes. When
we treated endothelial cells with EETs, all 4 isomers were
mitogenic in cerebral capillary endothelial cells, among
which 8,9-EET was the most potent and 5,6-EET was the
least potent. The dose-dependent relationship of 8,9-EET on
increasing 3H-thymidine incorporation is shown in Figure 5a.
VEGF treatment was used as a positive control of a known
angiogenic mitogen for endothelial cells21 (Figure 5a and 5b).
Both 8,9-EET and VEGF at their most effective concentrations significantly increased cell numbers (Figure 5b).
To begin to explore the pathway of EET-induced mitogenic effect, we examined 3H-thymidine incorporation before and
after treatment with an inhibitor of protein kinase C (PKC),
Myr␺PKC-I, at a concentration tested to inhibit PKC activity
previously,22 and an inhibitor of tyrosine kinase, genistein. As
can be seen in Figure 6, whereas Myr␺PKC-I did not inhibit
3
H-thymidine incorporation, genistein completely abolished
the effect of EETs, suggesting involvement of a tyrosine
kinase in EET-induced mitogenesis. The use of these inhibitors alone had no effect on the growth of endothelial cells
(Figure 6).
EET-Induced Morphogenesis of Endothelial Cells
on Matrigel
To further characterize the mitogenic and angiogenic effects
of EETs, we performed 2 sets of experiments. First, we plated
capillary endothelial cells on thin Matrigel with and without
EETs. Endothelial cells plated at 20 000 cells per chamber on
Matrigel in the presence of 8,9-EET in EBM-2 formed
capillary-like structure when examined at 18 hours after
seeding (Figure 7b). Similar capillary-like structures were
also induced by 11,12-EET and 14,15-EET (data not shown).
The same number of cells cultured in EBM-2 alone failed to
form long cordlike structures (Figure 7a). Capillary-like
structures were also formed in medium containing 5% FBS
and growth factors (MV medium) (see Materials and Methods and Figure 7c), which was used as a positive control. The
total length of tubes formed under different treatments was
measured. Tubes formed with EET treatment or with MV
medium were significantly longer than those with EBM-2
(Figure 7d).
Second, we performed a standard Matrigel plug implantation to examine whether EETs were able to induce angiogenesis in vivo. In this regard, we compared the number of
functional blood vessels in Matrigel plugs supplemented with
or without EETs. VEGF was used as a positive control. After
9 days of subcutaneous implantation, the Matrigel blocks
were recovered and examined. Matrigel formed a gel block
and was readily distinguished from surrounding tissue. Matrigel with vehicle produced little or no local reaction or
angiogenic response. To confirm that the observed growth of
vessels in Matrigel plugs was not artifactual, we sectioned
Matrigel plugs to look for functional vessels indicated by
linear structures containing red blood cells after hematoxylin
and eosin staining. As clearly seen in Figure 8, there was
marked angiogenesis in the plug treated with 200 nmol/L
8,9-EET (Figure 8a and 8b). The number of vessels in each
high-magnification view was counted. There were significantly more vessels in the Matrigel plugs treated with EET
than vehicle controls (Figure 8c).
Discussion
Involvement of astrocytes and astrocyte-produced EETs in
increasing 3H-thymidine incorporation and endothelial cell
morphogenesis was demonstrated in the present study with
the use of primary cultures of astrocytes and cerebral capillary endothelial cells. Our findings suggest that metabolites of
AA formed by P450 epoxygenase in astrocytes are involved
in angiogenesis in the central nervous system.
6
Stroke
December 2002
Downloaded from http://stroke.ahajournals.org/ by guest on June 18, 2017
Figure 7. Endothelial cell morphogenesis on
Matrigel. a, Cerebral capillary endothelial cells
cultured in serum-free EBM-2 on Matrigel for
18 hours. b, Cerebral capillary endothelial cells
cultured in serum-free EBM-2 medium with
addition of 100 nmol/L of 8,9-EET on Matrigel
for 18 hours form tubelike structures. c, Cerebral capillary endothelial cells cultured in MV
medium on Matrigel for 18 hours form tubelike
structures. d, The effect of EETs on endothelial
cell morphogenesis on Matrigel is quantitatively
expressed as the ratio between the total length
of tubes formed in EET-treated or MV medium
compared with that in EBM-2. EET treatment
significantly (*P⬍0.05) increased endothelial cell
tube formation on Matrigel. Data represent typical images from at least 3 independent experiments. Bar⫽0.2 mm.
Astrocytes have been shown to participate in angiogenesis
both in vitro and in vivo. Bovine retinal microvascular
endothelial cells formed capillary-like structures, which exhibited a positive DiI-Ac-LDL staining, on coculture with rat
brain astrocytes.19 When embryonic astrocytes were transplanted into the cerebral cortex of adult rats, capillary-like
structures were observed at the graft-host interface zone.23
This finding suggests that host endothelial cells migrate
toward the activated astrocytes and undergo differentiation to
capillaries. The observation, in the present study, that
capillary-like structures were only formed when astrocytes
were cocultured with endothelial cells supports the role of
astrocytes in angiogenesis. However, it must be noted that
cultured astrocytes proliferate readily, which is not typical of
astrocytes in vivo, although previous studies have shown that
astrocytes can divide during development in adults and
certainly under pathological conditions.24,25
We showed that astrocyte-conditioned medium was mitogenic on cerebral capillary endothelial cells. Inhibition of
P450 activity by 17-ODYA in astrocytes significantly reduced the conditioned medium–induced mitogenesis on brain
capillary endothelial cells, whereas blockade of P450 activity
in endothelial cells had no effect, indicating that mitogenic
factor(s) are released from astrocytes and that their formation
Figure 8. Angiogenic effect of EETs in
vivo by histological analysis of recovered
Matrigel plugs. a, Presence of vessels
(arrowheads) in Matrigel (MG) supplemented with EETs. Note a vessel
(arrows) extended into the Matrigel from
the connective tissue (CT). b, Highpowered micrograph of boxed area in
panel a showing a well-formed vessel
(arrows) with clear red blood cells and
endothelium lining. c, Quantitative analysis of numbers of vessels per ⫻40 view
in vehicle controls and EETsupplemented Matrigel blocks. Bar⫽50
␮m in a; bar⫽20 ␮m in b.
Zhang and Harder
Downloaded from http://stroke.ahajournals.org/ by guest on June 18, 2017
involves activation of astrocytic P450 epoxygenase. 17ODYA is a potent epoxygenase inhibitor (IC50⫽100 nmol/L).
It inhibits P-450 ␻-hydroxylase activity at higher concentrations. However, no study to date has shown any effect of
17-ODYA on the activity of non-P450 enzymes. The finding
that AA-induced increase in 3H-thymidine uptake was
blocked by nordihydroguaiaretic acid (NDGA) or low-dose
ketoconazole indicates that AA metabolites, and not AA
itself, act as mediators.26 Consistent with this notion, astrocytes express the epoxygenase P450 2C11, and astrocytes
readily metabolize AA into EETs8,27 and release EETs into
medium.28 Thus, our data in the present study suggest that a
P450 2C11– catalyzed production of EETs is involved in
astrocyte-induced mitogenesis.
Involvement of EETs in stimulating cell proliferation has
been demonstrated in other cell types. EETs have been shown
to promote cell proliferation in primary cultures of rabbit
proximal tube cells.29 It has been reported that 8,9- and
14,15-EETs stimulated thymidine incorporation when they
were administrated to mesangial cells in culture.26,30,31 On the
other hand, inhibitors of lipoxygenase (caffeic acid) or
cyclooxygenases (indomethacin) alone had no significant
effects on cell growth. Given the special anatomic location of
astrocytes between neurons and vascular endothelial cells, it
is very possible that astrocytes function as an intermediate
cell type to signal endothelial cells in responding to neural
activity. Certainly, EETs were not the only factors in the
astrocyte-conditioned medium that promoted mitogenesis, as
indicated by the finding that 17-ODYA cannot totally abolish
conditioned medium–induced mitogenic effect. However, our
data suggest that EETs increase proliferation of cerebral
endothelial cells with a magnitude similar to that of VEGF at
their physiological concentrations and, more importantly, that
EETs alone can induce differentiation of the endothelial cells
on Matrigel. The level of EET-induced increase in 3Hthymidine incorporation was approximately 2-fold under our
experimental conditions, which could be due to the lipid
nature of synthetic EETs susceptible to oxidative degradation
since studies using a more stable sulfonamide EET derivative
have shown a larger increase in 3H-thymidine incorporation
on renal epithelial cells.32 The average amount of 8,9-EET in
the astrocyte medium was 70 nmol/L, measured by liquid
chromatographic– electrospray ionization–mass spectrometry.28 The level of exogenous 8,9-EET (100 nmol/L) we used
is thus close to the actual level made by astrocytes. However,
the bioactivity of some EET degradation products, such as
dihydroxyeicosatrienoic acids, in our experimental settings
requires further study.
Extensive work has been performed on members of the
VEGF family in the last decade. To focus on VEGF with
respect to capillary angiogenesis in the brain is not to
preclude other growth factors that play a role in this process.
Release of VEGF from astrocytes occurs only during brain
development33 and in the presence of hypoxic and other
insults.34 –36 However, EET production in astrocytes increased
after glutamate treatment,14 and P450 epoxygenase inhibition
reduced the response of cerebral blood flow to neuronal
excitation.37,38 Indeed, there is a positive correlation between
neuronal activity and vessel density in the brain.39 Angiogen-
EETs in Endothelial Cell Morphogenesis
7
esis has been shown in adult rat brain after increased
metabolic demands.40 In addition, EETs have been characterized as an endothelium-derived hyperpolarizing factor to
dilate cerebral vasculature.41 In the central nervous system,
the major source of EETs is considered to be astrocytes, and
EETs are involved in vasodilation and functional hyperemia
to excitatory neurons.41,42 In vivo angiogenesis is accompanied by vasodilation. Agents capable of inducing vasodilation, such as endothelial nitric oxide synthase, played a
predominant role in VEGF-induced angiogenesis.43 We also
found that EETs augmented the effect of VEGF in our system
(preliminary data). Thus, EETs may represent an astrocytederived angiogenic mitogen in regulating physiological angiogenesis and may be involved in other growth factor–
mediated angiogenesis in the brain.
EETs may significantly contribute to the proangiogenic
program of capillary endothelial cells by triggering important
molecules in the signal transduction cascade such as tyrosine
kinase and mitogen-activated kinase.44,45 A PKC inhibitor
had no effect on EET-induced increase of 3H-thymidine
incorporation, indicating that PKC might not be involved.
Alternatively, the result could be a reflection of enhancing
EET incorporation by inhibition of PKC.46 Further experiments are required to reveal the downstream targets of EETs.
In summary, our findings provide direct evidence for a
novel mechanism involved in astrocyte-mediated angiogenesis and support a key role of astrocytes in regulating blood
flow to match neuronal metabolic demand. Further experiments should be undertaken to reveal the signal cascade
involved, which will provide more information for understanding the mechanism of angiogenesis in brain and developing treatment for diseases with disrupted balance of blood
vessel formation in the central nervous system.
Acknowledgments
This work was supported by National Institutes of Health/National
Heart, Lung, and Blood Institute grant PO1 HL 59996 to Dr Harder.
We thank Rachel Kraemer and Joan Anders for preparation and
maintenance of astrocyte cultures and Dr Meetha Medhora for
technical advice.
References
1. Araque A, Carmignoto G, Haydon PG. Dynamic signaling between
astrocytes and neurons. Annu Rev Physiol. 2001;63:795– 813.
2. Temburni MK, Jacob MH. New functions for glial in the brain. Proc Natl
Acad Sci U S A. 2001;98:3631–3632.
3. Harder DR, Zhang C, Gebremedhin D. Astrocytes function in matching
blood flow to metabolic activity. News Physiol Sci. 2002;17:27–31.
4. Ling T, Mitrofanis J, Stone J. Origin of retinal astrocytes in the rat:
evidence of migration from the optic nerve. J Comp Neurol. 1989;286:
345–352.
5. Chan-Ling T, Stone J. Factors determining the migration of astrocytes
into the developing retina: migration does not depend on intact axons or
patent vessels. J Comp Neurol. 1991;303:375–386.
6. Schnitzer J. Retinal astrocytes: their restriction to vascularized parts of the
mammalian retina. Neurosci Lett. 1987;78:29 –32.
7. Jiang B, Alibezhadian M, Caldwell RB. Astrocytes modulate retinal
vasculogenesis: effects on endothelial cell differentiation. Glia. 1995;
15:1–10.
8. Munzenmaier DH, Harder DR. Cerebral microvascular endothelial cell
tube formation: role of astrocytic epoxyeicosatrienoic acid release. Am J
Physiol. 2000;278:H1163–H1167.
9. Alkayed NJ, Narayanan J, Gebremedhin D, Medhora M, Roman RJ,
Harder DR. Molecular characterization of an arachidonic acid epoxygenase in rat brain astrocytes. Stroke. 1996;27:971–979.
8
Stroke
December 2002
Downloaded from http://stroke.ahajournals.org/ by guest on June 18, 2017
10. Capdevila J, Pramanik B, Napoli JL, Manna S, Falck JR. Arachidonic
acid epoxidation: epoxyeicosatrienoic acids are endogenous constituents
of rat liver. Arch Biochem Biophys. 1984;231:511–517.
11. Laniado-Schwartzman ML, Davis KL, McGiff JC, Levere RD, Abraham
NG. Purification and characterization of cytochrome P-450-dependent
arachidonic acid epoxygenase from human liver. J Biol Chem. 1988;263:
2536 –2542.
12. Murphy S, Pearce B, Jeremy J, Dandona P. Astrocytes as eicosanoidproducing cells. Glia. 1988;1:241–245.
13. Deleted in proof.
14. Alkayed NJ, Birks EK, Narayanan J, Petrie KA, Kohler-Cabot AE,
Harder DR. Role of P-450 arachidonic acid epoxygenase in the response
of cerebral blood flow to glutamate in rats. Stroke. 1997;28:1066 –1072.
15. Leffler CW, Fedinec AL. Newborn piglet cerebral microvascular
response to epoxyeicosatrienoic acids. Am J Physiol. 1997;273:
H333–H335.
16. Picciano PT, Johnson B, Walenga RW, Donovan M, Borman BJ, Douglas
WH, Kreutzer DL. Effects of D-valine on pulmonary artery endothelial
cell morphology and function in cell culture. Exp Cell Res. 1984;151:
134 –147.
17. Abbott NJ, Hughes CC, Revest PA, Greenwood J. Development and
characterization of a rat brain capillary endothelial culture: towards an in
vitro blood brain barrier. J Cell Sci. 1992;103:23–37.
18. Mccarthy KD, de Vellis J. Preparation of separate astroglial and oligodendroglial cell cultures from rat cerebral tissue. J Cell Biol. 1980;85:
890 –902.
19. Laterra J, Christopher G, Goldstein GW. Astrocytes induce neural microvascular endothelial cells to form capillary-like structures in vitro. J Cell
Physiol. 1990;144:204 –215.
20. Passaniti A. A simple, quantitative method for assessing angiogenesis and
antiangiogenic agents using reconstituted basement membrane, heparin,
and fibroblast growth factor. Lab Invest. 1992;67:519 –528.
21. Leung DW, Cachianes G, Kuang WJ, Goeddel DV, Ferrara N. Vascular
endothelial growth factor is a secreted angiogenic mitogen. Science.
1989;246:1306 –1309.
22. Lange A, Gebremedhin D, Narayanan J, Harder D. 20-Hydroxyeicosatetraenoic acid-induced vasoconstriction and inhibition of potassium
current in cerebral vascular smooth muscle is dependent on activation of
protein kinase C. J Biol Chem. 1997;272:27345–27352.
23. Suarez I, Bodega G, Rubio M, Garcia-Segura LM, Fernandez B.
Astroglial induction of in vivo angiogenesis. J Neural Transplant Plast.
1994;5:1–10.
24. O’Callaghan JP. Quantitative features of reactive gliosis following
toxicant-induced damage of the CNS. Ann N Y Acad Sci. 1993;679:
195–210.
25. Rakic P. Limits of neurogenesis in primates. Science. 1985;227:
1054 –1056.
26. Sellmayer A, Uedelhoven WM, Weber PC, Bonventre JV. Endogenous
non-cyclooxygenase metabolites of arachidonic acid modulate growth
and mRNA levels of immediate-early response genes in rat mesangial
cells. J Biol Chem. 1991;266:3800 –3807.
27. Amruthesh SC, Boerschel MF, McKinney JS, Willoughby KA, Ellis EF.
Metabolism of arachidonic acid to epoxyeicosatrienoic acids, hydroxyeicosatetraenoic acids, and prostaglandins in cultured rat hippocampal
astrocytes. J Neurochem. 1993;61:150 –159.
28. Nithipatikom K, Grall AJ, Holmes BB, Harder DR, Falck JR, Campbell
WB. Liquid chromatographic-electrospray ionization-mass spectrometric
analysis of cytochrome P450 metabolites of arachidonic acid. Anal
Biochem. 2001;298:327–336.
29. Burns KD, Capdevila J, Wei S, Breyer MD, Homma T, Harris RC. Role
of cytochrome P-450 epoxygenase metabolites in EGF signaling in renal
proximal tubule. Am J Physiol. 1995;269:C831–C840.
30. Harris RC, Homma T, Jacobson HR, Capdevila J. Epoxyeicosatrienoic
acids activate Na⫹/H⫹ exchange and are mitogenic in cultured rat glomerular mesangial cells. J Cell Physiol. 1990;144:429 – 437.
31. Homma T, Zhang JY, Shimizu T, Prakash C, Blair IA, Harris RC.
Cyclooxygenase-derived metabolites of 8,9-epoxyeicosatrienoic acid are
potent mitogens for cultured rat glomerular mesangial cells. Biochem
Biophys Res Commun. 1993;191:282–288.
32. Chen JK, Falck JR, Reddy KM, Capdevila J, Harris RC. Epoxyeicosatrienoic acids and their sulfonamide derivatives stimulate tyrosine phosphorylation and induce mitogenesis in renal epithelial cells. J Biol Chem.
1998;273:29254 –29261.
33. Ment LR, Stewart WB, Fronc R, Seashore C, Mahooti S, Scaramuzzino
D, Madri JA. Vascular endothelial growth factor mediates reactive angiogenesis in the developing brain. Brain Res Dev Brain Res. 1997;100:
52– 61.
34. Stone J, Ltin A, Gnessin H, Chan-Ling T, Pe’er Y, Keshet R. Normal
retinal vascularisation is mediated by the hypoxia-induced expression of
vascular endothelial growth factor by retinal glia. Exp Eye Res. 1994;
59:S65. Abstract.
35. Plate KH. Mechanism of angiogenesis in the brain. J Neuropathol Exp
Neurol. 1999;58:313–320.
36. Mart HH, Risau W. Systemic hypoxia changes the organ-specific distribution of vascular endothelial growth factor and its receptors. Proc Natl
Acad Sci U S A. 1998;95:15809 –15814.
37. Bhardwaj A, Northington FJ, Carhuapoma JR, Falck JR, Harder DR,
Traystman RJ, Koehler RC. P-450 epoxygenase and NO synthase inhibitors reduce cerebral blood flow response to N-methyl-D-aspartate. Am J
Physiol. 2000;279:H1616 –H1624.
38. Harder DR, Alkayed NJ, Lange AR, Gebremedhin D, Roman RJ. Functional hyperemia in the brain: hypothesis for astrocyte-derived vasodilator
metabolites. Stroke. 1998;29:229 –234.
39. Borowsky IW, Collins RC. Metabolic anatomy of brain: a comparison of
regional capillary density, glucose metabolism, and enzyme activities.
J Comp Neurol. 1989;288:401– 413.
40. Isaacs KR, Anderson BJ, Alcantara AA, Black JE, Greenough WT.
Exercise and the brain: angiogenesis in the adult rat cerebellum after
vigorous physical activity and motor skill learning. J Cereb Blood Flow
Metab. 1992;12:110 –119.
41. Campbell WB, Gebremedhin D, Pratt PF, Harder DR. Identification of
epoxyeicosatrienoic acids as endothelium-derived hyperpolarizing
factors. Circ Res. 1996;78:415– 423.
42. Harder DR, Roman RJ, Gebremedhin D. Molecular mechanisms controlling nutritive blood flow: role of cytochrome P450 enzymes. Acta
Physiol Scand. 2000;168:543–549.
43. Fukumura D, Gohongi T, Kadambi A, Izumi Y, Ang J, Yun CO, Buerk
DG, Huang PL, Jain RK. Predominant role of endothelial nitric oxide
synthase in vascular endothelial growth factor-induced angiogenesis and
vascular permeability. Proc Natl Acad Sci U S A. 2001;98:2604 –2609.
44. Chen JK, Wang DW, Falck JR, Capdevila J, Harris RC. Transfection of
an active cytochrome P450 arachidonic acid epoxygenase indicates that
14,15-epoxyeicosatrienoic acid functions as an intracellular second messenger in response to epidermal growth factor. J Biol Chem. 1999;274:
4764 – 4769.
45. Chen JK, Capdevila J, Harris RC. Overexpression of C-terminal Src
kinase blocks 14,15-epoxyeicosatrienoic acid-induced tyrosine phosphorylation and mitogenesis. J Biol Chem. 2000;275:13789 –13792.
46. Shivachar AC, Willoughby KA, Ellis EF. Effect of protein kinase C
modulation on 14,15-epoxyeicosatrienoic acid incorporation into
astroglial phospholipids. J Neurochem. 1995;65:338 –346.
Cerebral Capillary Endothelial Cell Mitogenesis and Morphogenesis Induced by
Astrocytic Epoxyeicosatrienoic Acid
Chenyang Zhang and David R. Harder
Downloaded from http://stroke.ahajournals.org/ by guest on June 18, 2017
Stroke. published online October 17, 2002;
Stroke is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231
Copyright © 2002 American Heart Association, Inc. All rights reserved.
Print ISSN: 0039-2499. Online ISSN: 1524-4628
The online version of this article, along with updated information and services, is located on the
World Wide Web at:
http://stroke.ahajournals.org/content/early/2002/10/17/01.STR.0000037787.07479.9A.citation
Permissions: Requests for permissions to reproduce figures, tables, or portions of articles originally published
in Stroke can be obtained via RightsLink, a service of the Copyright Clearance Center, not the Editorial Office.
Once the online version of the published article for which permission is being requested is located, click
Request Permissions in the middle column of the Web page under Services. Further information about this
process is available in the Permissions and Rights Question and Answer document.
Reprints: Information about reprints can be found online at:
http://www.lww.com/reprints
Subscriptions: Information about subscribing to Stroke is online at:
http://stroke.ahajournals.org//subscriptions/