Abstract
purpose. To investigate the hypothesis that nitric oxide (NO) in the optic nerve
heads of rats with chronic moderately elevated intraocular pressure
(IOP) contributes to neurotoxicity of the retinal ganglion cells, the
presence of the three isoforms of nitric oxide synthase (NOS) was
determined in the tissue.
methods. Unilateral chronic moderately elevated IOP was produced in rats by
cautery of three episcleral vessels. Histologic sections of optic
nerves from eyes with normal IOP and with chronic moderately elevated
IOP were studied by immunohistochemistry and by immunoblot analysis.
Polyclonal antibodies to NOS-1, NOS-2, NOS-3, and glial fibrillary
acidic protein (GFAP) were localized with immunoperoxidase.
results. In the optic nerve of rat eyes with normal IOP, NOS-1 was
constitutively present in astrocytes, pericytes and nerve terminals in
the walls of the central artery. NOS-2 was not present in eyes with
normal IOP. In these eyes, NOS-3 was constitutively present in the
vascular endothelia of large and small vessels. Rat eyes treated with
three-vessel cautery had sustained elevated IOP (1.6 fold) for at least
3 months. In these eyes, no obvious changes in NOS-1 or NOS-3 were
noted. However, at time points as early as 4 days of chronic moderately
elevated IOP, NOS-2 appeared in astrocytes in the optic nerve heads of
these eyes and persisted for up to 3 months. Immunoblot analysis did
not detect differences in NOS isoforms.
conclusions. The cellular distributions of constitutive NOS isoforms in the rat
optic nerve suggest physiological roles for NO in this tissue. NOS-1 in
astrocytes may produce NO as a mediator between neighboring cells. NO,
produced by NOS-1 in pericytes and nitrergic nerve terminals and by
NOS-3 in vascular endothelia, is probably a physiological vasodilator
in this tissue. In eyes with chronic moderately elevated IOP, NOS-2 is
apparently induced in astrocytes. The excessive NO production that is
associated with this isoform may contribute to the neurotoxicity of the
retinal ganglion cells in eyes with chronic moderately elevated
IOP.
Although studies of patients with glaucoma indicate that elevated
intraocular pressure (IOP) is the single most common finding in
glaucoma,
1 the pathophysiological mechanisms by which
elevated IOP leads to cellular events that are neurodestructive in the
optic nerve head are unknown. Knowledge of cellular mediator pathways
that contribute to glaucomatous optic neuropathy may lead to new
therapeutic approaches that slow or prevent further damage to the axons
of the retinal ganglion cells. Achieving pharmacologic neuroprotection
for the treatment of glaucoma may augment or even supersede the current
clinical goal of lowering IOP.
Nitric oxide (NO), a dissolved gas with a half-life of a few
seconds,
2 is an important mediator with diverse
physiological roles and certain pathologic roles in various tissues,
including the central nervous system (CNS) and the
eye.
3 4 5 6 7 In the CNS, NO has been implicated in
neurodegenerative diseases such as stroke, Alzheimer’s disease,
multiple sclerosis, and amyotropic lateral sclerosis. In animal models
of neurodegeneration, NO can be both neuroprotective and
neurodestructive.
Nitric oxide is formed from
l-arginine by nitric oxide
synthase (NOS). Molecular cloning has identified three distinct genes
expressing NOS isoforms: neuronal NOS (nNOS or NOS-1), endothelial NOS
(eNOS or NOS-3) and inducible NOS (iNOS or NOS-2). The nomenclature was
derived from the tissue in which they were first studied, and the
numbering was derived from the order in which they were
cloned.
8 NOS-1 and NOS-3 are constitutive, are present
physiologically, and are calcium dependent. Under conditions of
degeneration and inflammation, both isoforms may be
upregulated.
5 In contrast, NOS-2 is not constitutive, is
Ca
2+ independent, and is induced after immunologic
challenge and neuronal injury.
5 9
In the rat CNS, all three NOS isoforms have been well characterized
with immunohistochemistry.
10 11 12 13 In the retina, NOS-1 has
been identified in amacrine cells, nerve fibers of the inner and outer
plexiform layers, the ganglion cell layer, photoreceptor ellipsoids,
bipolar cells, and large choroidal vessels.
10 14 15 However, the optic nerve of normal rats, although not studied in great
detail, is reported to be devoid of NOS.
10
NOS-1 mediates early neuronal injury after middle cerebral artery
occlusion
16 and is blocked pharmacologically with an
NOS-1–selective inhibitor 7-nitroindazole.
17 NOS-1 null
transgenic mice sustain reduced infarct volumes, compared with
age-matched control animals, after permanent focal middle cerebral
artery occlusion.
18 Blockage of neurotoxicity with NOS
inhibitors in primary cortical neuronal cultures has also been
demonstrated.
5 In contrast, NOS-3, a major regulator of
vascular hemodynamics and blood vessel relaxation, is neuroprotective.
Transgenic mice without NOS-3 experience increased infarct volumes
after middle cerebral artery occlusion.
19 Inducible NOS,
which is not normally expressed in the brain, can be induced in
astrocytes and microglia after viral infection or trauma. Induction of
NOS-2 produces large quantities of NO for sustained periods and results
in neuronal damage.
20 21 22 NOS-2 may contribute to
neurodegeneration in several human diseases.
23
Our laboratory has recently reported the presence of the constitutive
isoforms NOS-1 and NOS-3 in the normal human optic nerve head and has
demonstrated all three isoforms of NOS in the optic nerve head of
patients with primary open-angle glaucoma, suggesting induction of
NOS-2 in glaucoma.
24 Excessive NO, synthesized by NOS-1
and/or NOS-2, may be neurodestructive in the optic nerve heads of
patients with glaucoma.
Given the findings of NOS isoforms in glaucomatous optic nerve heads
and the implication of NO as a neurodestructive agent, we have studied
the presence, localization and distribution of the three NOS isoforms
in the rat optic nerve. We chose the rat because this species provides
a potentially useful model for studying optic neuropathy caused by
chronic moderately elevated IOP
25 for the eventual purpose
of screening promising pharmacologic compounds.
26 We
confirmed the elevated IOP and retinal ganglion cell loss in this model
and, using commercially available antibodies for immunohistochemistry
and immunoblot analysis, we demonstrated NOS isoforms in the anterior
optic nerve of normal rats and rats with chronic moderately elevated
IOP. Our results on the NOS-2 isoform are similar to those that we have
previously reported for the human optic nerve head.
All experiments were performed in accordance with the ARVO
Statement for the Use of Animals in Ophthalmic and Vision Research.
AlbinoWistar rats (200–300 g) were fed ad libitum and maintained in
temperature-controlled rooms on a 12-hour light–12-hour dark cycle
(light period from 6 AM to 6 PM). Animals were anesthetized by
intraperitoneal injection of a mixture of xylazine (12 mg/kg) and
ketamine (80 mg/kg). IOP was measured with a pneumotonometer (Mentor 30
Classic; BioRad, Santa Ana, CA) under light anesthesia, as previously
described. IOP measurements were begun within 5 minutes of inducing
anesthesia, and the animals were awake within 30 minutes of the IOP
measurement. Rat eyes were anesthetized with one drop of topical 0.5%
proparacaine HCl (Bausch & Lomb Pharmaceuticals, Tampa, FL). The mean
IOP of four to six consecutive measurements was recorded. IOP was
measured before surgery, within 2 hours after surgery, and once or
twice a week for up to 3 months after surgery. All IOP measurements
were made at the same time of day (10 AM–12 noon) by one of three
observers, and the measurements were completed within 5 minutes.
Unilateral, three-vessel cautery for elevating IOP
25 was
performed as previously described.
26 Briefly, three of the
four to five major trunks formed by limbal-derived veins were exposed
at the equator of the eye by incising the conjunctiva. Each vein was
lifted with a small muscle hook or forceps and cauterized by direct
application of an ophthalmic, disposable cautery (Model RS201; Roboz,
Rockville, MD) against the muscle hook. Care was taken to avoid
cauterizing sclera directly. Immediate retraction and absence of
bleeding of the cauterized ends of the vessels were noted as successful
cauterization. All animals had unilateral chronic moderately elevated
IOP and the contralateral eye was either untouched or subjected to a
sham operation. For each animal, the contralateral eye always served as
the comparative control. On a given day, mean ± SD was derived
for all control and surgical (three-vessel cautery) eyes. Significant
differences between surgical and control eyes were determined byχ
2 analysis with Student’s
t-test
for independent means for each day on which measurements were
performed.
For immunohistochemistry, after 4, 14, 28, 45, and 84 days of
unilateral chronic moderately elevated IOP, groups of four to six
animals at each time point were killed with an overdose of the
anesthesia mixture described earlier. Enucleation of both eyes was
performed with a suture placed at 12 o’clock for proper orientation,
and eyes were fixed in fresh 4% paraformaldehyde for 1 hour.
Subsequently, the cornea, iris, lens, and vitreous were removed. The
posterior segment was fixed in fresh 4% paraformaldehyde for 1 hour
and then transferred to 70% ethanol overnight. Fixed tissue was washed
in 0.2% glycine in phosphate-buffered saline (PBS; pH 7.4), embedded
in paraffin, and oriented for 6-μm sagittal sections. Five sections
per optic nerve were examined by use of immunohistochemistry, with one
of the sections serving as a negative control.
The isoforms of NOS were identified using a polyclonal antibody to
human NOS-1, which recognizes an epitope at the amino terminus (working
dilution, 1:300), a polyclonal antibody to human NOS-3, which
recognizes an epitope at the carboxyl terminus (working dilution,
1:500), and a polyclonal antibody to mouse NOS-2, which recognizes an
epitope at the carboxyl terminus (working dilution 1:25; all polyclonal
antibodies to human NOS from Santa Cruz Biotechnology, Santa Cruz, CA).
To identify glial cells, a monoclonal antibody (Sigma, St. Louis,
MO) to glial fibrillary acidic protein (GFAP) was used (working
dilution, 1:400). Slides were preincubated with 5% skim milk for 30
minutes and then incubated with primary antibody overnight at 4°C.
Primary antibodies were localized by immunoperoxidase staining
using commercially available reagents (Vector, Burlingame, CA). The
biotinylated secondary antibody was incubated on the sections for 30
minutes, washed with PBS, and reacted with the streptavidin-peroxidase
conjugate for 30 minutes. After washing, sections were incubated with
the substrate mixture (1.5 mg 3,3-diaminobenzidine tetrahydrochloride
and 50 μl 30% hydrogen peroxide in 0.1 mM Tris phosphate [pH
7.6]). The sections were reacted until brown staining appeared, in
approximately 1 to 3 minutes, washed in PBS, counterstained with
hematoxylin, and dehydrated. A coverslip was mounted on the slide with
sealer (Cytoseal 60; Fisher Scientific, Pittsburgh, PA).
Representative sections of all samples were stained simultaneously to
control variation in the reactions. Negative controls were performed by
eliminating primary antibody from the incubation medium, using the
antibody preabsorbed with the appropriate antigenic peptide, or by
replacing the primary antibody with nonimmune serum followed by
immunoperoxidase staining. Slides were examined by microscope (BH2;
Olympus, Tokyo, Japan), and images were recorded by digital photography
and stored as a computer file.
For immunoblot analysis, 12 animals were killed after 28 days of
unilateral chronic moderately elevated IOP. Immunoblots were performed
on lysates of individual rat optic nerves that were dissected from
surrounding tissues. Protein (10 μg) was loaded and run on to 10%
Tris-glycine gels (Novex, San Diego, CA) and then transferred to a
nitrocellulose membrane (Hybond-c; Amersham, Arlington Heights, IL).
Membranes were blocked with 2% milk and treated with 0.5% Tween-20
before application of the primary antibodies (working dilution,
1:1000). After 1 hour, the membranes were washed, and secondary
antibody with conjugated horseradish peroxidase (Santa Cruz)
was applied (working dilution, 1:10,000). After further washing,
the blots were developed with enhanced chemiluminescence western blot
detection reagents (Amersham).
Figure 2 demonstrates the cellular localizations of NOS-1 and NOS-3 in the optic
nerve head of the rat. In the anterior portion of the optic nerve head,
corresponding to the neck region, and in the transition area,
containing cells organized transversely across the nerve bundles,
NOS-1–positive, granular staining appeared to be distributed
throughout the cytoplasm of individual cells
(Fig. 2A) . The pattern of
distribution of NOS-1–positive cells was similar to the pattern of
cells that were positive for GFAP, a cytoskeletal marker for astrocytes
(Fig. 2B) . However, many more cells were positive for GFAP than for
NOS-1, suggesting that not all astrocytes contain NOS-1. In the glial
columns that were posterior to the transition area, most cells were
positive for NOS-1
(Fig. 2C) .
Occasionally, cells in the transition zone, closely wrapped around a
capillary wall and abutting the nerve fiber bundles, were positive for
NOS-1
(Fig. 2D) . This NOS-1–positive immunoreactivity was not
localized to the endothelial cells of the capillary and did not
correspond to GFAP staining (compare to
Fig. 2B ). This pattern suggests
that NOS-1 may be present in pericytes in the rat optic nerve head.
In the optic nerve head near the vitreal surface, staining for NOS-1
was also associated with the arterial vasculature, with a punctatelike
pattern in the outer walls of the vessel that was suggestive of nerve
terminals
(Fig. 2E) . This type of staining was not apparent around
capillaries or veins. Of note, the endothelial cells bordering the
vascular lumen were not positive for NOS-1.
Eyes with chronic moderately elevated IOP were evaluated for
NOS-1–positive cells on days 4, 14, 28, and 45. No difference in the
density of positive cells could be discerned by inspection.
Figure 2 also demonstrates the presence of cells that were positive for
NOS-3. These cells were localized primarily to the vasculature.
NOS-3–positive cells are present in both the arterial and venous
systems in the optic nerve head
(Fig. 2G) . The staining pattern was
diffuse throughout the cell body and localized to the endothelial cells
with no evidence of NOS-3 in the surrounding perivascular cells.
Throughout the optic nerve, including the neck region, transition zone
and the area of myelination, numerous capillaries were present, with
positive NOS-3 immunoreactivity in the endothelial cells
(Fig. 2H) .
Eyes with chronic moderately elevated IOP for 4, 14, 28, and 45 days
were evaluated for NOS-3–positive cells. No difference in the density
of positive cells could be discerned by inspection.
NOS-2–positive cells were present in the optic nerve heads of rat eyes
with chronic moderately elevated IOP but were not present in
contralateral control eyes with normal IOP
(Fig. 3) . Examination at low power of optic nerve head tissue from rat eyes
with 4 days of chronic moderately elevated IOP demonstrated clusters of
cells in the transition zone that were positive for NOS-2
(Fig. 3A) .
Not all cells were positive for NOS-2, and not all regions contained
NOS-2–positive cells. At high power, NOS-2–positive cells had the
distribution of GFAP-positive cells or astrocytes
(Fig. 3B) . At 28 and
84 days of chronic moderately elevated IOP, NOS-2–positive cells,
presumably astrocytes, were identified in the transition region of
optic nerve heads (
Figs. 3C 3E , respectively) but not in contralateral
control eyes with normal IOP
(Figs. 3D 3F) . Usually, several cells
could be identified per tissue section. NOS-2–positive cells were
always in locations associated with glia. The vascular endothelia of
some capillaries may have been NOS-2–positive in these eyes. Cells
that were positive for NOS-2 were not present in the optic nerve proper
or the retina. No loss of nerve fibers or disorganization of the optic
nerve head tissue was histologically apparent after 3 months of chronic
moderately elevated IOP.
Control slides, obtained by omitting the primary antibody or by
preincubating the primary antibody with specific antigenic peptides,
were all negative for specific NOS staining.
Figure 2F demonstrates the
absence of specific staining when the primary antibody for NOS-1 was
omitted. Additional figures demonstrating the absence of specific
staining under control conditions are not presented. However, because
we can demonstrate different cellular localizations for the NOS
isoforms in the rat optic nerve, comparisons can be made between the
figures presented for each primary antibody used for each isoform to
demonstrate the specificity of any one primary antibody.