Abstract
purpose. The administration of glucocorticoids as topical or systemic
medications may lead to the development of ocular hypertension through
the induction of morphologic and biochemical changes in the trabecular
meshwork leading to a reduction in the facility of aqueous outflow.
Glucocorticoids exert their physiological effects by binding to and
activating glucocorticoid and mineralocorticoid receptors. The activity
of glucocorticoids is critically regulated at a prereceptor level by
the two isozymes of 11β-hydroxysteroid dehydrogenase. The purpose of
this study was to determine the distribution of glucocorticoid target
receptors and the isozymes of 11β-hydroxysteroid dehydrogenase (11β
-HSD) that regulate the activity of glucocorticoids at a prereceptor
level in human and rat ocular tissues.
methods. Horizontal sections of normal adult human and rat eyes were cut and
hybridized with 35S-labeled cRNA probes specific for the
glucocorticoid receptor, mineralocorticoid receptor, and 11β-HSD
types 1 and 2 using in situ hybridization. Immunohistochemical analysis
of glucocorticoid and mineralocorticoid receptors using monoclonal
antibodies was carried out on rat eye tissue sections. Whole rat eyes
were homogenized and the activity of 11β-HSD types 1 and 2 in the eye
assessed as the percentage conversion of tritiated corticosterone to
tritiated 11-dehydrocorticosterone when corticosterone was added to the
homogenate.
results. In the rat ocular tissues mRNAs encoding glucocorticoid receptor,
mineralocorticoid receptor, and 11β-HSD types 1 and 2 were detected
in nonpigmented ciliary epithelium, trabecular meshwork, corneal
epithelium and endothelium, and anterior lens epithelium.
Immunohistochemistry confirmed the presence of glucocorticoid and
mineralocorticoid receptors at these sites. Activity of both isozymes
of 11β-HSD was demonstrated in homogenized rat eyes (percentage
conversion of tritiated corticosterone to 11-dehydrocorticosterone;
mean ± SD, 11β-HSD 1 = 15% ± 5.3%, 11β-HSD 2 =
7.9% ± 2.8%). In both human and rat eyes, expression of mRNAs
encoding glucocorticoid receptor and 11β-HSD type 1 was high in the
trabecular meshwork and lens epithelium, whereas expression of mRNAs
encoding the mineralocorticoid receptor and 11β-HSD type 2 was high
in nonpigmented ciliary epithelium and corneal epithelium and
endothelium.
conclusions. Glucocorticoid target receptors and the enzymes regulating
glucocorticoid activity at these receptors are present in mammalian
ocular tissues, which regulate aqueous humor formation and outflow.
Alteration in the number or affinity of receptors or in the activity of
regulatory enzymes may alter the susceptibility of certain individuals
to the effects of glucocorticoids on intraocular
pressure.
Topical or systemic administration of glucocorticoids (GCs)
produces a rise in intraocular pressure in a proportion of the normal
population
1 by decreasing the facility of aqueous
outflow.
2 If this pressure rise is sustained it may result
in optic disc cupping and visual field loss similar to that seen in
primary open-angle glaucoma (POAG). Glucocorticoids may cause a
reduction in aqueous outflow through the many cellular and morphologic
changes that they induce in trabecular meshwork (TM) cells. These
changes include alterations in extracellular matrix
production,
3 4 5 6 cell size,
7 8 9 nuclear
size
7 and DNA content,
7 cytoskeletal
organization,
8 9 phagocytic activity,
10 and
protease activity.
11 In addition, dexamethasone has been
shown to alter Na–K–Cl cotransport,
12 an effect that, by
altering TM cell volume, may affect the facility of aqueous outflow in
intact TM. Glucocorticoid responsiveness occurs with a far higher
prevalence among POAG patients than among normal subjects, with over
90% of POAG patients being considered GC responders,
13 compared with 30% to 35% of the normal population. Steroid responders
have been reported to be at increased risk of developing POAG compared
with nonresponders,
14 and there are numerous reports of
raised intraocular pressure in patients with Cushing’s
syndrome,
15 a collection of clinical signs and symptoms
caused by chronically elevated levels of endogenous GCs.
Electron microscopic studies comparing postmortem eyes of POAG and
GC-induced glaucoma cases have reported similar but not identical
morphologic changes in the TM
16 (increased accumulation of
extracellular material). POAG specimens typically show so-called
sheath-derived plaques in the cribriform layer of the TM and beneath
the inner wall endothelium of Schlemm’s canal.
17 Specimens from cases of corticosteroid glaucoma typically show
accumulation of a fingerprint-like–arranged basement membrane material
in the cribriform and outer corneoscleral regions and also a fine
fibrillar material beneath the inner wall endothelium of Schlemm’s
canal.
16
Glucocorticoids exert their effects by binding to intracellular
receptors of two types, glucocorticoid (GR) and mineralocorticoid (MR)
receptors. The ligand-receptor complex then migrates to the nucleus and
binds to specific DNA sequences called glucocorticoid response elements
(GRE), altering the transcription of target genes.
18 This,
in turn, leads to altered synthesis of proteins. Glucocorticoid
receptors have previously been demonstrated in human TM,
19 whereas MRs have been shown to be present in rabbit,
20 bovine,
21 and human
22 23 ocular tissues. In
addition Starka et al. have previously demonstrated effects of cortisol
and aldosterone on ionic composition of the lens and aqueous
humor.
24 Recently, an additional level of control of GC
action has been identified. The isozymes of 11β-hydroxysteroid
dehydrogenase (11β-HSD) act as important regulators of GC activity at
a prereceptor level.
25 26 27 These isozymes catalyze the
interconversion of cortisol, the major glucocorticoid in the human, to
its inactive metabolite cortisone (corticosterone to
11-dehydrocorticosterone in the rat). In vivo, the type 1 isozyme acts
primarily as a reductase, converting inert cortisone to active
cortisol, thereby increasing the access of cortisol to
GR.
25 The type 2 isozyme acts primarily as a high affinity
dehydrogenase in vivo, converting active cortisol to inactive cortisone
and thereby conferring aldosterone specificity on intrinsically
nonselective MRs in aldosterone target tissues such as the distal
nephron of the kidney.
26 27 Deficiency of 11β-HSD type 2
results in the syndrome of apparent mineralocorticoid excess
(SAME)
28 29 in which cortisol illicitly activates MRs,
causing sodium retention, systemic hypertension, and hypokalemia. These
processes may also occur in other salt-transporting epithelia. We
investigated, using in situ hybridization and immunohistochemistry, the
distribution of GR and MR and 11β-HSD types 1 and 2 in human and rat
ocular tissues. We also investigated the activity of 11β-HSD types 1
and 2 in homogenized rat ocular tissues using enzyme activity assays.
Rat cDNA clones for GRs,
30 MRs,
31 and
11β-HSD type 1
32 and type 2
33 were
linearized using the appropriate restriction enzymes. Antisense and
sense complementary RNA probes were synthesized from the resulting
templates using the appropriate RNA polymerases and[α
-
35S]UTP (>1000 Ci/mM; Amersham Life
Science, Little Chalfont, Buckinghamshire, UK). Human cDNA clones for
GR,
34 MR,
34 11β-HSD type 1
35 and type 2
36 were linearized using the appropriate
restriction enzymes. Antisense and sense complementary RNA probes were
synthesized from the resultant templates using the appropriate RNA
polymerases. Probes were purified on Nick columns (Pharmacia Biotech,
Uppsala, Sweden) and checked for size and purity on denaturing
polyacrylamide gels.
All procedures used in these studies followed the tenets of the
Declaration of Helsinki and the ARVO Statement for the Use of Animals
in Ophthalmic and Vision Research. Informed consent was obtained before
obtaining human tissue samples. Full ethical approval was granted for
all portions of this study by the Lothian Health Research Ethics
subcommittee and by the Western General Hospital NHS Trust Research and
Development Ethics Committee. Rat eyes were obtained from healthy adult
male Lister-Hooded rats and paraffin-embedded. Human eyes (obtained
from Glaucoma Research Foundation, San Francisco, CA, and the Queen
Victoria Hospital, East Grinstead West Sussex, UK) were cut in lateral
and medial parasagittal planes, and sections were paraffin-embedded.
Horizontal sections (5-μm-thick) were cut using a microtome (Leitz
GmBH, Wetzlar, Germany) and sections placed on
3-aminopropyltriethoxysilane (APES 2%; Sigma, St. Louis, MO)–coated
slides. Sections were deparaffinized by immersion in histoclear (2 × 10 mins; Fisher Scientific, Loughborough, Leicestershire, UK).
Histoclear was removed by washing in ethanol (100% × 2 minutes;
Merck, Poole, UK). Sections were rehydrated by immersion in graded
alcohols (100%, 100%, 95%, 85%, 70%, 50%, 30% ethanol). Ethanol
was removed by washing in sodium chloride (0.9%). This was followed by
immersion in Triton-X (0.3%; Koch Light, Suffolk, UK) in 1×
phosphate-buffered saline (1× PBS for 15 minutes) after which sections
were washed twice in 1× PBS (5 minutes). Tissue sections were then
digested in trizma–HCl (100 mM, pH 8; Sigma), EDTA (50 mM; Sigma)
containing proteinase K (30 minutes, 37°C; Sigma), then washed in
glycine (0.1%; Merck) in 1× PBS. Sections were then postfixed in
paraformaldehyde (4%; Fisher Scientific), washed in 1× PBS (2 ×
5 minutes) followed by acetylation in acetic anhydride (0.25%; Sigma)
in triethanolamine (0.1 M, pH 8; Sigma), washed in 1× PBS (1 × 3
minutes), dehydrated in graded alcohols, and air-dried. Sections were
incubated with prehybridization buffer made up of diethylpyrocarbonate
water, sodium chloride (5 M), trizma base (1 M), 50× Denhardt’s
(Sigma), salmon testes DNA (Sigma), EDTA (250 mM; Sigma), and yeast
tRNA (GIBCO–BRL Products, Paisley, UK) in deionized formamide
(50°C × 2 hours; Sigma). Hybridization was then carried out by
incubation with 35S-labeled riboprobe (1 ×
106 cpm) in hybridization buffer containing
diethylpyrocarbonate water, sodium chloride (5 M), trizma base (1 M),
50× Denhardt’s, salmon testes DNA, EDTA (250 mM), and yeast tRNA in
deionized formamide (50°C × 16 hours). After hybridization,
sections were washed in SSC (15 minutes) and incubated with RNase A
(100 μg/ml, 37°C for 1 hour; Sigma). Sections were then washed to
increasing stringencies to a maximum of 0.1× SSC (60°C for 1 hour).
After dehydration through graded alcohols, sections were placed against
hyperfilm βmax (2 weeks at 4°C; Amersham) and
autoradiographs developed. After this, sections were dipped in
photographic emulsion (NTB-2; Kodak, Rochester, NY) and exposed
(4°C for 3 weeks) before being developed and counterstained with
hematoxylin and eosin (Sigma).
Areas of specific mRNA expression on tissue sections were
identified by the appearance of silver grains. Grain counting (SEE-Scan
Image Analysis Systems UK) was performed on antisense and corresponding
sense sections. Background counts (counts from sense sections) were
subtracted from antisense counts and results expressed as multiples of
background. Positive control sections were used as follows: rat
hippocampus (GR, MR and 11β-HSD type 1), rat kidney (11β-HSD type
2), human cerebellum (GR), human liver (11β-HSD type 1), and human
kidney (MR and 11β-HSD type 2).
Whole rat eyes were removed and placed in a 10% solution of 40%
formaldehyde in phosphate buffer (pH 7) for 48 hours. Eyes were then
processed through graded alcohols (70%, 80%, 100% × 3; 90 minutes
each) and xylene (2 × 90 minutes; Merck Ltd.) and embedded in
paraffin wax. Tissue sections (5-μm-thick) were cut by microtome
(Leitz) and placed onto APES-coated tissue slides. Slide sections were
deparaffinized using xylene, which was removed by immersing sections in
absolute ethanol. Sections were rehydrated by immersion in ethanol
(70%) followed by water. Slides were then immersed in hydrogen
peroxide (3%; AAH Pharmaceuticals, Huddersfield, UK) to block
endogenous peroxidase and washed in water.
Endogenous biotin was blocked using a biotin blocking kit (Vector
Laboratories, Burlingame, CA). Monoclonal antibodies, Mab-7 (1/1000)
against GR
37 (provided courtesy of Kjel Fuxe,
Karolinska Institute, Sweden) and MR-4 (1/1000) against
MR
38 (provided courtesy of Zygmunt S. Krozowski, Baker
Medical Institute, Prahan, Victoria, Australia), were applied to tissue
sections and unbound antibody removed by washing with PBS (0.05 M).
Sections were next treated with biotin-labeled secondary goat antibody,
with unbound antibody again being removed by washing in PBS (0.05 M; 30
minutes at 22°C). A streptavidin/biotinylated peroxidase complex
(DAKO Ltd., Cambridge, UK) was added as a tertiary agent and bound to
biotin on the secondary antibody. Unbound avidin was removed by washing
in PBS (0.05 M). Hydrogen peroxide (3%) was added to the sections
along with the potential dye diaminobenzidine (DAB; Sigma). Tissue
sections were counterstained using Mayer’s hematoxylin. Sections were
then dehydrated by immersion in graded ethanols (70%, 80%, 100%) and
finally immersed in xylene and coverslipped using DPX mountant.
The activity of 11β-HSD types 1 and 2 was investigated by
measuring the ability of homogenized rat eyes to convert tritiated
corticosterone to tritiated 11-dehydrocorticosterone in the presence of
the essential cofactors NADP and NAD
+ for
11β-HSD types 1 and 2, respectively. Pooled tissues from 10 rat eyes
were homogenized in Krebs’ buffer (250 μl; without bovine serum
albumin [BSA]) in 3 × 10 second bursts in an Ystral Homogenizer
(Scientific Instrument Center, Liverpool, UK) and assayed for
protein colorimetrically (Biorad Laboratories, Hemel Hempstead,
Herts, UK) using the Lowry method.
39 40 Incubations were performed in triplicate at 37°C in 250 μl
containing 1.12 × 10
−8 M[
1,2,6,7-
3H
4]-corticosterone
(Amersham Life Science; 84 Ci/mM), ethanol 1% vol/vol, BSA 0.2 g/dl,
and tissue homogenate at 500 μg protein/ml and in the presence or
absence of NAD
+ or NADP
+ (2
mM; Sigma). The protein concentration was chosen to be in the linear
part of the relationship before the assessment of enzyme activity.
Separation was achieved using a C
18 Microbondapak
Column (30 cm; Millipore Waters, Watford, UK), using a mobile phase of
methanol:water (65:35) at a flow rate of 3.9 ml/min. Radiodetection was
carried out by online liquid scintillation counting (Optiflow, 3.5
ml/min; Berthold, Berlin, Germany) and results presented as
percentage of 11-dehydrocorticosterone to total of
11-dehydrocorticosterone and corticosterone. Incubations were also
carried out in the absence of protein and cofactor to allow the
subtraction of conversion under these conditions.
The results of studies on human eyes are shown in
Table 1 and
Figures 1 and 2 . In human eyes the expression of GR mRNA
(Fig. 1) was present in
nonpigmented ciliary epithelium, TM, lens epithelium, corneal
epithelium, and corneal endothelium. Expression was highest in lens
epithelium, TM, and nonpigmented ciliary epithelium. Glucocorticoid
receptor mRNA was also noted to a lesser extent in corneal epithelium
and endothelium. Expression of MR mRNA
(Fig. 2) was highest in
nonpigmented ciliary epithelium, corneal endothelium, and corneal
epithelium. Expression was also found, although to a lesser extent, in
lens epithelium and TM. 11β-HSD type 1 mRNA was expressed mainly in
TM, lens epithelium, and in nonpigmented ciliary epithelium
(Fig. 1) .
Expression was also detected in corneal epithelium and endothelium.
11β-HSD type 2 mRNA
(Fig. 2) was expressed most highly in
nonpigmented ciliary epithelium and in corneal epithelium. Expression
was also present in corneal endothelium, TM, and lens epithelium.
Control sections showed positive expression of MR and 11β-HSD type 2
(kidney), GR (cerebellum), and 11β-HSD type 1 (liver). In human eyes,
mRNA expression for GR, MR, and 11β-HSDs was not detected in iris
stroma, corneal stroma, or sclera.
Immunohistochemistry.
Immunohistochemical studies on rat eye sections confirmed the presence
of the protein product of GR- and MR-specific mRNAs in lens epithelium,
nonpigmented ciliary epithelium, and TM
(Fig. 3) . Mineralocorticoid receptor was expressed to a lesser extent than GR
in keeping with the findings of in situ hybridization.
Enzyme Activity Assay.
Activity of 11β-HSD types 1 and 2 was confirmed in homogenized rat
eyes. Activity was expressed as percentage conversion of tritiated
corticosterone to tritiated 11-dehydrocorticosterone
(Fig. 4) .
In Situ Hybridization.
The results of studies on rat eyes are shown in
Table 2 and
Figure 5 . In rat eyes, the expression of mRNAs encoding GR, MR, and 11β-HSD
types 1 and 2 was present in the lens, nonpigmented ciliary epithelium,
and TM. Glucocorticoid receptor mRNA was expressed highly in cells of
the TM, anterior lens epithelium, corneal epithelium, and nonpigmented
ciliary epithelium. Mineralocorticoid receptor mRNA was expressed less
highly than GR, and expression was highest in nonpigmented ciliary
epithelium, corneal epithelium, and corneal endothelium. Expression of
11β-HSD type 1 mRNA was also lower than GR and was most noticeable in
anterior lens epithelium, corneal endothelium, and tissues of the TM.
11β-HSD type 2 mRNA was expressed to a lesser extent than the type 1
isozyme, and expression was greatest in corneal endothelium,
nonpigmented ciliary epithelium, and TM
(Table 2) . Expression of mRNAs
for GR, MR, and 11β-HSDs was not detected in iris stroma, corneal
stroma, ciliary muscle, or sclera.
This study investigated the distribution of glucocorticoid target
receptors (GR and MR) and enzymes, which regulate glucocorticoid
hormone activity at these receptors (11β-HSD types 1 and 2) in human
and rat ocular tissues. In this study we demonstrated expression of GR,
MR, and 11β-HSD types 1 and 2 in lens, TM, ciliary epithelium, and
corneal epithelium and endothelium. The pattern of mRNA expression was
similar in human and rat ocular tissues. We also noted that MR and
11β-HSD type 2 mRNA expression colocalized in nonpigmented ciliary
epithelium and cornea, whereas GR mRNA expression colocalized with that
of 11β-HSD type 1 in lens and TM. We have therefore demonstrated
glucocorticoid target receptors (GR and MR) and regulatory enzymes
(11β-HSD types 1 and 2) at sites in ocular tissues responsible for
both the secretion and outflow of aqueous humor. This suggests that
these tissues may be potential sites of action for steroid hormones
with glucocorticoid and mineralocorticoid activity. Previous work by
Weinreb and coauthors
19 demonstrated the presence of
functional GRs in cultured human TM cells. We have now demonstrated the
presence of GR in lens epithelium and in nonpigmented ciliary
epithelium in addition to TM of human and rat eyes. These findings are
of interest given the well-documented effects of glucocorticoids on
intraocular pressure and their ability to induce cataract. The access
of GCs to GR in the brain and liver is regulated by the action of
11β-HSD type 1.
41 We have now shown 11β-HSD type 1
mRNA to be present in cornea, nonpigmented ciliary epithelium, ciliary
muscle, lens, and TM of human and rat eyes. This colocalization of GR
and 11β-HSD type 1 suggests that the ocular tissue is another site in
the body that has the capacity to regulate the activity of GCs at a
prereceptor level and corresponds with previous observations suggesting
a functional relationship between GR and 11β-HSD type 1 in brain and
kidney.
41 42 Hormones possessing mineralocorticoid
activity act through MR to influence ion and fluid transport mechanisms
in many cell types.
43 Ion and fluid transport mechanisms
play an important role in several key processes in the eye, including
secretion of aqueous humor
44 and maintenance of corneal
transparency.
45 Previous immunohistochemical studies have
demonstrated the presence of MR in ocular tissues, including ciliary
epithelium, cornea, and lens.
22 23 Further studies by
Schwartz and Wysocki
20 and Starka et al.
21 have identified aldosterone binding sites in mammalian ocular tissues,
including nonpigmented ciliary epithelium and lens. Starka et al. also
demonstrated the ability of aldosterone to alter the ionic composition
of the lens and aqueous humor.
24 In addition our studies
have demonstrated the presence of MR in the TM and cornea of rat and
human eyes. Mineralocorticoid receptor has been shown to bind both
cortisol and aldosterone. In the distal nephron, in vivo, the presence
of 11β-HSD type 2 ensures aldosterone-selective access to nonspecific
MR. 11β-HSD type 2 has not been reported previously in the eye. We
have demonstrated this isozyme in cornea, nonpigmented ciliary
epithelium, lens, and TM of human and rat eyes. The colocalization of
MR and 11β-HSD type 2 may indicate selective access for aldosterone
to MR in ocular tissues.
There is a considerable body of evidence to suggest a role for
glucocorticoids in the pathogenesis of disorders of intraocular
pressure (e.g., POAG). Glucocorticoid-induced cellular and morphologic
changes in TM may affect resistance to aqueous humor outflow and lead
to ocular hypertension and similar, although not identical,
ultrastructural changes have been found in the TM of patients with both
corticosteroid-induced glaucoma and POAG.
16 17 The effects
of GCs on ocular tissues in disease states such as POAG may be the
result of higher circulating levels of GCs in the blood and aqueous
humor,
46 with a previous report showing an ocular
hypotensive response in rabbits to RU486, a steroid receptor
antagonist.
47
However, alterations in the number or affinity of target receptors (GR
and MR) or indeed the activity of the 11β-HSDs will also alter the
effects of GCs at a cellular level (e.g., steroid-resistant
asthma,
48 hypertension,
49 and apparent
mineralocorticoid excess
28 29 ) and, thus, alter the impact
of GCs on intraocular pressure. Mutations in the 11β-HSD type 2 gene
cause impairment in renal MR function, and changes in hepatic 11β-HSD
type 1 function have been correlated with obesity.
50 It
remains to be investigated whether or not alterations in these crucial
enzymes occur in disorders of intraocular pressure.
The authors thank the Glaucoma Research Foundation of San
Francisco, California, and the Queen Elizabeth Hospital East Grinstead
(Sussex, UK) for providing human eye tissue and James Ironside
and William Shade for their assistance.